A Novel Approach to Neural Network-based Motion Cueing Algorithm for a Driving Simulator

Abstract

Generating realistic motion in a motion-based (dynamic) driving simulator is challenging due to the limited workspace of the motion system of the simulator compared to the motion range of the simulated vehicle. Motion Cueing Algorithms (MCAs) render accelerations by controlling the motion system of the simulators to provide the driver with a realistic driving experience. Commonly used methods such as Classical Washout-based MCA (CW-MCA) typically achieves suboptimal results due to scaling and filtering, which results in an inefficient usage of the workspace. The Model Predictive Control-based MCA (MPC-MCA) has been shown to achieve superior results and more efficient workspace use. However, it's performance is in practice constrained due to the computationally expensive operations and the requirement of an accurate prediction of future vehicle states. Finally, the Optimal Control (OC) has been shown to provide optimal cueing in an open-loop setup wherein the precalculated control signals are re-played to the driver. However, OC cannot be used in real-time with the driver-in-the-loop. Our work introduces a novel Neural Network-based MCA (NN-MCA), which is trained to imitate the behavior of the OC. After training, the NN-MCA provides an approximated model of the OC, which can run in real-time with the driver in-the-loop, while achieving similar quality. The experiments demonstrate the potential of this approach through objective evaluations of the generated motion-cues on the simulator model and the real simulator. A demonstration video for the performance comparison of the CW-MCA, Optimal-Control-based MCA (OC-MCA) and our proposed method is available at http://go.tum.de/708350.

A Review of Dynamic Vehicular Motion Simulators: Systems and Algorithms

“…This paper categorized these robotic machines based on their working principles and the number of DoF that they provide to facilitate smooth translational and rotational positioning in the 3d space. Furthermore, the Adaptive washout lter [6], [18][21] Optimal washout lter [22] OpDA algorithm [23] Sliding mode-based cueing [24][26], [31] Model predictive control [27] [29] Time-varying model predictive control [30], [32] Nonlinear model predictive control [33] Neural network [34] Fuzzy control system…”

“…In [33] (https://www.in.tum.de/leadmin/w00bws/i06/ Personal_Files/Emec_Ercelik/PMS_Final.mp4), the model was trained in an end-to-end manner, meaning that the neural network learned the complete next state of the simulator in addition to the control input signals.…”

The Impact of Physical Motion Cues on Driver Braking Performance: A Clinical Study Using Driving Simulator and Eye Tracker

“…To enhance simulator fidelity, efforts have been made in this research field to provide a more realistic sense of presence [ 14 ]. Thus, advanced artificial intelligence (AI) techniques, including deep neural network (DNN) [ 15 ], fuzzy logic [ 16 , 17 , 18 ], or genetic algorithm [ 19 ], have been exploited to optimize platform motion cueing in a high degrees-of-freedom (DOF) in the roll, pitch, and yaw axis. However, other studies underlined the high-cost issue of such developed software and hardware of motion platforms and intended to reduce simulators’ cost by decreasing the freedom to 3-DOF [ 20 ], 2-DOF [ 21 ], or even to completely static simulators [ 22 , 23 , 24 , 25 , 26 ].…”

Handiski simulator performance under PSO-based washout and control parameters optimization

“…The motion cueing algorithm, commonly called the washout filter, objective is to provide the user with the most accurate perception of sliding/piloting so that the user cannot distinguish the real situation from that simulated by using a set of translational and angular movements. The MCA generates a trajectory that tends to maximize human sensation while respecting the simulator's physical limits and returns the simulator platform back to its neutral position over time [10][11][12]. The classic approach algorithm was developed for the first time by [13,14].…”