Kalman filters applied to actively controlled railway vehicle suspensions

Abstract

This paper studies the development and application of Kalman filters for actively controlled railway vehicle suspensions. The basic requirement of state estimation for active suspensions of railway vehicles is discussed and two distinct track features that affect the design of a state observer are highlighted. Two effective methods to accommodate the different track characteristics based on two detailed case studies are presented in the paper. The principles described, however, apply to other uses of model-based estimation on railway vehicles. In the first study a state observer for the plan-view model of a two-axle vehicle is developed, primarily for the active steering of railway wheelsets. The observer using Kalman filters is formulated not only to estimate all necessary vehicle states, but also to include and hence estimate as part of the state variables the curve radius and cant of the railway track on which the vehicle is travelling. In the second study, Kalman filters are used to estimate state variables of the side-view model of a quarter car. The output is then used to implement the (active) vertical suspension using the principle of ‘skyhook’ damping in order to optimize the trade-off between the random and deterministic input requirements.