MIMO Vibration Control for a Flexible Rail Car Body: Design and Experimental Validation

Abstract

Car bodies of modern rail vehicles are designed as lightweight structures with the aim to minimize mass and thus operational energy demand. The central structural design requirements are given by the main static and dynamic loads. However, ride comfort becomes an increasingly important issue because the softer, more compliant structure exhibits low eigenfrequencies that significantly affect perceived passenger ride comfort. Various approaches have been taken to reduce comfort-relevant vibrations of the car body that can be grouped into vibration isolation and vibration damping approaches. The isolation approaches include passive, semi-active, and active concepts to decouple the car body from the bogeys and wheel sets. The active approaches are more complex and can affect the safety against derailment, but potentially lead to improved isolation performance over their passive counterparts (see Foo & Goodall (2000) and Stribersky et al. (1998)). The complementary vibration damping approaches intend to increase the elastic eigenmodes’ damping ratios. A passive approach is taken in Hansson et al. (2004), active control schemes have been proposed in Kamada et al. (2005) as well as Schandl et al. (2007) and Benatzky (2006). The latter two references treat the same metro configuration and actuation concept as the present work. This chapter presents LQG and weighted H2 MIMO control design methods for the vibration control of lightweight rail car body structures. These designs are studied and compared to achieve vibration reduction and passenger ride comfort improvement in a highly flexible metro rail car body. The metro car body structure is directly actuated via locally mounted Piezo stack actuators. Utilizing strain measurement signals, the control law actuates the structure with the aim of minimizing ride comfort-relevant acceleration signals across the car body interior. This system is subject to variations in damping and frequency of the flexible modes which pose a challenge for control design; the two control methods are implemented in a simulation as well as in a scaled experimental setup and their capabilities are investigated. The work is embedded in a rich series of research and publications treating various aspects of control design for flexible rail car bodies. A robust H∞-optimal control approach is surveyed in Kozek et al. (2011), including an overview on modeling, control design, simulation, identification, and experimental results. These research topics are focused on and detailed in Benatzky & Kozek (2005; 2007a;b); Benatzky et al. (2006; 2007); Bilik et al. (2006); 15