Flow: A Modular Learning Framework for Mixed Autonomy Traffic

Abstract

The rapid development of autonomous vehicles (AVs) holds vast potential for transportation systems through improved safety, efficiency, and access to mobility. However, the progression of these impacts, as AVs are adopted, is not well understood. Numerous technical challenges arise from the goal of analyzing the partial adoption of autonomy: partial control and observation, multi-vehicle interactions, and the sheer variety of scenarios represented by real-world networks. To shed light into near-term AV impacts, this article studies the suitability of deep reinforcement learning (RL) for overcoming these challenges in a low AV-adoption regime. A modular learning framework is presented, which leverages deep RL to address complex traffic dynamics. Modules are composed to capture common traffic phenomena (stop-and-go traffic jams, lane changing, intersections). Learned control laws are found to improve upon human driving performance, in terms of system-level velocity, by up to 57% with only 4-7% adoption of AVs. Furthermore, in single-lane traffic, a small neural network control law with only local observation is found to eliminate stop-and-go traffic - surpassing all known model-based controllers to achieve near-optimal performance - and generalize to out-of-distribution traffic densities.

17 pages, 8 figures, 5 tables. 2021 IEEE Transactions on Robotics (T-RO)

References in corpus (8)

• Dissipation of stop-and-go waves via control of autonomous vehicles: Field experiments
• OpenAI Gym
• Delays, Inaccuracies and Anticipation in Microscopic Traffic Models
• CoLight: Learning Network-level Cooperation for Traffic Signal Control
• Gated Feedback Recurrent Neural Networks
• Smoothing Traffic Flow via Control of Autonomous Vehicles
• The Intelligent Driver Model with Stochasticity -- New Insights Into Traffic Flow Oscillations
• Simulation to Scaled City: Zero-Shot Policy Transfer for Traffic Control via Autonomous Vehicles

Cited by in corpus (13)

• DeeP-LCC: Data-EnablEd Predictive Leading Cruise Control in Mixed Traffic Flow
• Unified Automatic Control of Vehicular Systems with Reinforcement Learning
• Safety-Critical Traffic Control by Connected Automated Vehicles
• Traffic expertise meets residual RL: Knowledge-informed model-based residual reinforcement learning for CAV trajectory control
• System-size dependence of a jam-absorption driving strategy to remove traffic jam caused by a sag under the presence of traffic instability
• Learning to Control and Coordinate Mixed Traffic Through Robot Vehicles at Complex and Unsignalized Intersections
• Autonomous Driving Small-Scale Cars: A Survey of Recent Development
• Cooperation for Scalable Supervision of Autonomy in Mixed Traffic
• AD4RL: Autonomous Driving Benchmarks for Offline Reinforcement Learning with Value-based Dataset
• Learning to Control DC Motor for Micromobility in Real Time with Reinforcement Learning
• Lessons in Cooperation: A Qualitative Analysis of Driver Sentiments towards Real-Time Advisory Systems from a Driving Simulator User Study
• Cooperative Advisory Residual Policies for Congestion Mitigation
• A Fairness-Oriented Multi-Objective Reinforcement Learning approach for Autonomous Intersection Management