[Paper Review] An Integrated Framework of Decision Making and Motion Planning for Autonomous Vehicles Considering Social Behaviors

Peng Hang, Chen Lv+4 more|arXiv (Cornell University)|May 22, 2020

Traffic control and management35 references4 citations

TL;DR

This paper proposes an integrated decision-making and motion planning framework for autonomous vehicles that models social behaviors of surrounding traffic participants using Stackelberg Game theory and potential fields. By combining non-cooperative game theory with Model Predictive Control (MPC), the approach enables safe, adaptive lane changes under diverse social interaction scenarios, demonstrating effectiveness in three validation scenarios with improved decision-making and trajectory planning accuracy.

ABSTRACT

This paper presents a novel integrated approach to deal with the decision making and motion planning for lane-change maneuvers of autonomous vehicle (AV) considering social behaviors of surrounding traffic occupants. Reflected by driving styles and intentions of surrounding vehicles, the social behaviors are taken into consideration during the modelling process. Then, the Stackelberg Game theory is applied to solve the decision-making, which is formulated as a non-cooperative game problem. Besides, potential field is adopted in the motion planning model, which uses different potential functions to describe surrounding vehicles with different behaviors and road constrains. Then, Model Predictive Control (MPC) is utilized to predict the state and trajectory of the autonomous vehicle. Finally, the decision-making and motion planning is then integrated into a constrained multi-objective optimization problem. Three testing scenarios considering different social behaviors of surrounding vehicles are carried out to validate the performance of the proposed approach. Testing results show that the integrated approach is able to address different social interactions with other traffic participants, and make proper and safe decisions and planning for autonomous vehicles, demonstrating its feasibility and effectiveness.

Motivation & Objective

• To address the challenge of safe and effective lane-change maneuvers in autonomous vehicles by modeling social interactions with surrounding traffic.
• To integrate decision-making and motion planning into a unified framework that accounts for diverse driving styles and intentions of surrounding vehicles.
• To improve the adaptability and safety of autonomous vehicles in complex traffic environments through a constrained multi-objective optimization approach.
• To validate the framework under realistic scenarios involving varying social behaviors of surrounding traffic participants.

Proposed method

• Formulates the decision-making process as a non-cooperative Stackelberg Game problem to model strategic interactions between the autonomous vehicle and surrounding traffic.
• Employs potential fields with behavior-specific functions to represent surrounding vehicles and road constraints in motion planning.
• Utilizes Model Predictive Control (MPC) to predict the autonomous vehicle’s future states and trajectories over a finite horizon.
• Integrates decision-making and motion planning into a single constrained multi-objective optimization problem to ensure consistency and safety.
• Uses different potential functions to encode distinct social behaviors (e.g., aggressive, cautious) of surrounding vehicles.
• Validates the framework using three test scenarios with varying social interaction patterns to assess performance and robustness.

Experimental results

Research questions

• RQ1 How can social behaviors of surrounding vehicles be effectively modeled to improve decision-making in autonomous lane changes?
• RQ2 What is the impact of integrating decision-making and motion planning on the safety and adaptability of autonomous vehicles?
• RQ3 How does the proposed framework handle diverse driving styles and intentions of surrounding traffic participants?
• RQ4 Can the integrated approach maintain safety and performance across multiple realistic traffic interaction scenarios?
• RQ5 What role does Stackelberg Game theory play in enabling strategic decision-making under uncertainty?

Key findings

• The integrated framework successfully handles diverse social interaction patterns, including cooperative and competitive behaviors, during lane-change maneuvers.
• The use of Stackelberg Game theory enables the autonomous vehicle to anticipate and react to surrounding vehicles’ intentions in a strategic manner.
• Model Predictive Control ensures accurate trajectory prediction and real-time adaptability under dynamic conditions.
• The framework demonstrates improved safety and decision quality across all three test scenarios, with no collision or unsafe maneuvers observed.
• The integration of behavior-specific potential fields enhances motion planning robustness by reflecting different social behaviors in the environment model.
• The constrained multi-objective optimization formulation ensures consistent and feasible solutions across both decision-making and motion planning components.