Motion Cueing Algorithms for Full-Motion Simulations in Low-Gravity Environments

Abstract

Motion simulation is used in the development and training of novel vehicle control concepts, such as for future lunar landings. To ensure the transferability of the results, the vehicle motions must be reproduced as realistically as possible. Since reduced gravity cannot be accurately simulated on Earth, this poses a challenge for scenarios in low-gravity environments, and appropriate error-handling strategies are to be defined. This work implements two motion cueing algorithms for such a low-gravity environment simulation, which convert the vehicle motion into appropriate movements of a robot-based motion platform. A classical washout filter in cylindrical coordinates aims at exploiting the rotational degree of freedom of this simulator type, and an optimal motion cueing algorithm attempts to minimize the perception error by taking into account human perception models that account for different gravitational forces. Both algorithms strive for correct representation of angular rates and linear accelerations while assuming an orientation error. To compensate for this deficiency, a force feedback stick is integrated into the motion cueing to represent the current vehicle orientation. A software-in-the-loop analysis evaluates both algorithms concerning working space utilization and compares the expected motion perceptions with two perception models - a simple transfer function-based model and a state-observer-based approach. In a preliminary piloted simulator campaign, the algorithms are further tested in combination with force feedback. The results indicate that both algorithms are valid approaches with potential for further optimization through modified parameterization. The force feedback stick improves the situational awareness and is positively perceived