Answers are grounded in this paper and cite their sources.
Reliable operation of rail vehicles with independently rotating wheels (IRWs) requires explicit control strategies that remain effective under degraded sensing conditions, as these systems do not provide inherent passive lateral self-stabilisation. While high-performance control architectures typically rely on extensive sensor suites, including inertial measurement units, wheel encoders and displacement sensors combined via sensor fusion, such configurations are vulnerable to sensor degradation or complete failure. This paper investigates the minimum sensor requirements that still allow for stable lateral state reconstruction and control under degraded sensing conditions. The primary objective is to ensure safe and stable lateral guidance with reduced sensing capabilities rather than nominal full-sensing operation. Based on a detailed vehicle model, an observability analysis is conducted to identify sensor configurations that preserve reconstructability of the relevant lateral and yaw dynamics without relying on inertial or optical measurements. The results show that encoder measurements at the front and rear wheel carriers constitute a minimal and practically robust fallback configuration. Based on this setup, two complementary fallback control strategies are proposed. The first is a virtual wheelset approach that enforces a kinematic coupling between left and right wheel speeds, thereby reproducing the stabilising effect of conventional rigid wheelsets. The second is an observer-based control scheme that reconstructs lateral vehicle states using only wheel speed measurements from the identified fallback sensors. Both approaches are evaluated in simulation and validated experimentally on a 1:5 scale IRW test vehicle. The results demonstrate that stable lateral guidance can be maintained under severe sensor degradation using only minimal wheel speed information, enabling robust fallback operation without inertial or absolute position measurements.
No takes yet. Share an insight, caveat, or question.
Tobias Posielek (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: