Summary of Race-sm: Reinforcement Learning Based Autonomous Control For Social On-ramp Merging, by Jordan Poots
RACE-SM: Reinforcement Learning Based Autonomous Control for Social On-Ramp Merging
by Jordan Poots
First submitted to arxiv on: 5 Mar 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Machine Learning (cs.LG); Robotics (cs.RO)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes a novel learning-based approach to autonomous vehicle control, specifically for parallel-style on-ramp merging. The existing non-learning based solutions rely on rules and optimization, which present significant challenges. Recent advancements in Deep Reinforcement Learning have shown promise, but existing learning-based approaches often ignore other highway vehicles and rely on inaccurate road traffic assumptions. To address this issue, the proposed model explicitly considers the utility to both the ego vehicle and its surrounding vehicles, which may be cooperative or uncooperative. The reward function is designed using Social Value Orientation, weighing the vehicle’s level of social cooperation. Simulation results demonstrate the importance of considering surrounding vehicles in reward function design and show that the proposed model matches or surpasses existing approaches in terms of collisions while introducing socially courteous behavior. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper tries to make autonomous vehicles behave better when merging onto a highway. Right now, most solutions use rules and optimization, but they don’t always work well. Some newer ideas using Deep Reinforcement Learning have shown promise, but they often ignore other cars on the road. This new approach takes into account how all the cars around you might be behaving. It uses something called Social Value Orientation to decide what’s best for everyone involved. The results show that this approach can help avoid accidents and make the roads safer. |
Keywords
* Artificial intelligence * Optimization * Reinforcement learning
