Safe Learning Reference Governor: Theory and Application to Fuel Truck Rollover Avoidance

TL;DR

This paper introduces a learning reference governor (LRG) method that enforces constraints and improves control performance over time, demonstrated on fuel truck rollover prevention with liquid sloshing effects.

Contribution

The paper develops a novel LRG approach capable of learning and enforcing constraints in systems with unknown models, applied specifically to fuel truck rollover avoidance.

Findings

LRG effectively prevents fuel truck rollovers in simulations

The approach improves command tracking while maintaining safety constraints

Applicable to systems with black-box models or direct hardware learning

Abstract

This paper proposes a learning reference governor (LRG) approach to enforce state and control constraints in systems for which an accurate model is unavailable, and this approach enables the reference governor to gradually improve command tracking performance through learning while enforcing the constraints during learning and after learning is completed. The learning can be performed either on a black-box type model of the system or directly on the hardware. After introducing the LRG algorithm and outlining its theoretical properties, this paper investigates LRG application to fuel truck (tank truck) rollover avoidance. Through simulations based on a fuel truck model that accounts for liquid fuel sloshing effects, we show that the proposed LRG can effectively protect fuel trucks from rollover accidents under various operating conditions.