Robust system state estimation for active suspension control in high-speed tilting trains

The interaction between the railway vehicle body roll and lateral dynamics substantially influences the tilting system performance in high-speed tilting trains, which results in a potential poor ride comfort and high risk of motion sickness. Integrating active lateral secondary suspension into the tilting control system is one of the solutions to provide a remedy to roll–lateral interaction. It improves the design trade-off for the local tilt control (based only upon local vehicle measurements) between straight track ride comfort and curving performance. Advanced system state estimation technology can be applied to further enhance the system performance, i.e. by using the estimated vehicle body lateral acceleration (relative to the track) and true cant deficiency in the configuration of the tilt and lateral active suspension controllers, thus to further attenuate the system dynamics coupling. Robust H∞ filtering is investigated in this paper aiming to offer a robust estimation (i.e. estimation in the presence of uncertainty) for the required variables, In particular, it can minimise the maximum estimation error and thus be more robust to system parametric uncertainty. Simulation results illustrate the effectiveness of the proposed schemes.