抄録
In this paper we present a method of improving actuator control performance of a rapid prototyping bogie (RPB), which has 9 electric actuators. The actuators emulate missing railway bogie suspension components and are used to determine an optimal parameter set for them via a test bench system. In the original RPB hybrid control system, the actuators were force-controlled independently of each other, according to the behaviour of a virtual suspension model. During the RPB performance tests deterioration in the control performance was observed, which was caused by dynamic coupling between the multiple actuators. To solve this issue, we have developed a new controller based upon the dynamically substructured systems (DSS) method, which ensures accurate synchronisation of numerical and physical states at their common boundaries. A description of this DSS approach applied to the RPB problem therefore forms the core of this paper. To confirm the effectiveness of the new approach, we implemented a DSS-based controller into an existing proof-of-concept test rig, which has characteristics that are common to the RPB - in particular, it has multiple actuators attached to the same rigid body component and the existence of rubber bushes to protect the actuators from excessive bending moments. A simplified, linear model of a single railway wheel-set system was chosen as an example. Each actuator emulates not only a virtual primary suspension but also virtual creep forces. A state-space DSS approach was adopted to improve numerical conditioning. Through random excitation tests with the DSS controller, we confirmed that the actuators were well synchronized with their numerical models and that the previously observed and unexpected frequency components were removed. As a consequence, the DSS approach was determined as a viable framework for future research on the RPB system.
