Closing the loop in a treadmill-based 2- wheeler simulator: Development of a Communication Bridge and PI-Controller for a Treadmill

Abstract

This thesis project aims to further develop an existing simulator that uses a largescale treadmill from Rodby AB as a driving platform to simulate a real driving environment for 2-wheeled vehicles (e-scooter). To keep the vehicle centered on the treadmill a rope is attached to the vehicle, which introduces unwanted interference. Additionally, a motion capture system from Qualisys AB is used to track the movement and position of the rider. This serves as a controlled environment to study behaviors of riders on 2-wheeled vehicles, such as riding under the influence of alcohol. For a more realistic simulation, a closed-loop control system that adjusts the speed of the treadmill based on the position of the e-scooter was implemented, replacing the rope. The objective of the project was to stabilize a 2-wheeled vehicle at the center of the treadmill without the use of physical restraints. Position measurement was implemented using high precision Qualisys optical motion capture system that transmits position information in the 6 Degrees of Freedom (6DoF) in real time to a custom control system built in Python. A custom Proportional-Integral (PI) control logic was implemented to calculate and control the speed of the treadmill depending on the riders position. It was able to adjust the speed of the treadmill depending on the riders position but was not fast enough to keep the rider centered on the treadmill when the rider accelerated aggressively, due to the treadmills limited acceleration capabilities compared to the e-scooter. In addition to software latency which further slows the system. Further development is required for the control system to operate accurately to simulate a real driving environment for 2-wheeled vehicles.