Coupling multibody dynamics and hydraulic actuators for indirect Kalman filtering and real-time simulation

Abstract

The computer simulation of complex mechanical systems can be performed using multibody dynamics. Multibody models often include a description of hydraulic actuators, which can introduce numerical stiffness into the models. This numerical stiffness problem can be alleviated by properly selecting either the multibody model or the friction model of a hydraulic cylinder. Furthermore, the actual dynamic state of such models can be inferred using information-fusing techniques such as nonlinear Kalman filters, where information from a physical system can be combined with its multibody model. This synergy can result in advanced simulation-based solutions. Additionally, the real-time capabilities of coupled multibody and hydraulic models can be utilized in product development, user training, and other product processes.

The objective of this study is to couple multibody dynamics and hydraulic actuators for indirect Kalman filtering and real-time simulation in the framework of heavy machinery. In this regard, efficient formulations for the coupled simulation of multibody and hydraulic models are developed here in a monolithic framework. The simulation models are further combined with an indirect Kalman filter to estimate actual dynamic states, thereby, representing a real system. Furthermore, the application of the real-time capabilities for coupled multibody and hydraulic models are demonstrated in the areas of user training and product development. This study incorporates both academic and industrial examples of hydraulically actuated systems. The results demonstrate the accuracy, efficiency, utility, and wide applicability for the developed simulation-based solutions.