Hybrid control approach improves safety and efficiency of autonomous trucks
Learned Terminal Costs for Short-Horizon Model Predictive Control of Autonomous Trucks in Interactive Highway Traffic
Machine LearningArtificial IntelligenceRobotics
Summary
Driving big trucks autonomously on highways is hard because they are heavy, large, and need to interact safely and smoothly with other cars. The paper describes a way to combine two different methods: one that carefully plans short-term moves with strict safety rules, and another that learns about long-term driving from experience. The researchers use the learned long-term information to guide the short-term plans, helping the truck make safer and better decisions quickly. Tests in simulations show that their combined method avoids crashes completely and performs better than using either method alone.
What this means in practice
- •For autonomous vehicle engineers: Develop control systems for highway trucks that safely navigate interactive traffic using short planning horizons boosted with learned long-term costs.
- •For traffic simulation developers: Incorporate improved hybrid control models for simulating safer and more realistic heavy truck behavior in highway traffic scenarios.
Authors
Hooman Hamze, Benjamin Houshmand, Deepthi Pathare, Erik Börve, Morteza Haghir Chehreghani, Nikolce Murgovski
Abstract
Autonomous highway driving for heavy-duty trucks is a challenging control problem. The greater mass, dimensions and intricate dynamics increase the requirements for cooperation from surrounding traffic in critical maneuvers, requiring a careful balance between safety constraints and efficient long-horizon decision making. Model Predictive Control (MPC) provides rigorous constraint satisfaction but relies on terminal cost which can limit its effectiveness at short, computationally affordable horizons. Reinforcement Learning (RL) captures long-term driving behavior but offers no constraint guarantees and tends toward overly conservative policies when trained under sparse collision penalties. In this paper, we propose a hybrid MPC and RL framework for highway driving in autonomous trucks, where a finite-horizon MPC, solved via dynamic programming, uses a pre-trained RL model for terminal cost estimation. We evaluate the framework in a simulation environment with reactive surrounding traffic and compare it against pure MPC with zero terminal cost, rollout with terminal cost estimation using a base policy, and pure RL. Results show that the hybrid approach achieves zero collisions with the lowest cost of any configuration.