A Framework for Motion Cueing Algorithm Development Demonstrated With a Nonlinear MPC-Based Approach

Kurzfassung

Motion Cueing Algorithms (MCAs) convert simulated vehicle movements into commands for physical motion simulators, enabling the reproduction of realistic motion sensations for a pilot. Various MCA approaches have been developed over the past few decades, ranging from classic washout filters to optimization-based formulations and machine learning-based methods. While the MCA is typically the primary focus of development, significant effort is often devoted to implementing the associated infrastructure, including the interface to the motion simulator, reference trajectory generation, data processing, and visualization. This additional work is time-consuming and distracts from the core task of algorithm development. In this article, we present a modular framework that facilitates the development, implementation, and evaluation of MCAs on motion simulators. The framework provides standardized interfaces, integrated evaluation tools, and real-time communication with the motion simulator, thereby simplifying the overall integration effort and establishing a foundation for future motion cueing research and experimental studies. The framework is demonstrated using a nonlinear model predictive control-based MCA on the DLR Robotic Motion Simulator with a flight dynamics model of a Diamond DA42.