Vehicle Sideslip Angle Estimation Using Deep Reinforcement Learning Combined with Unscented Kalman Filter
Liguang Wu, Wei Wang, Penghui Li, Yueying Zhu, Enrico Meli · 2025
Read Vehicle Sideslip Angle Estimation Using Deep Reinforcement Learning Combined with Unscented Kalman Filter by Liguang Wu, Wei Wang, Penghui Li, Yueying Zhu, Enrico Meli (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12736524.txt.
Description
A research team pairs a classic filter with a learning agent to estimate a car's sideways drift angle, in a 2025 paper. The sideslip angle is a core quantity for stability control and active safety. Kalman filters estimate it well, but their noise settings are fixed by hand and degrade on tricky roads.
The team's method lets a deep reinforcement learning agent tune the filter's noise matrices on the fly. The agent watches vehicle motion and filter performance, then adjusts the settings to cut estimation error. A policy optimization algorithm trained the agent across varied speeds, road grip levels, and sensor noise. Root-mean-square error fell by over thirty percent versus the fixed filter. Accuracy held up across the tested conditions. The authors suggest the hybrid approach could extend to other vehicle states that safety systems need.
Catalog record
| CREATOR | Liguang Wu, Wei Wang, Penghui Li, Yueying Zhu, Enrico Meli |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |