In the following master’s thesis the design, implementation, and experimental evaluation of an electric wheelchair simulator that integrates dynamic modeling, real-time control and a three-dimensional virtual environment, is presented. The simulator combines a physics-based suspension model with a Stewart platform motion system, enabling realistic reproduction of the electric wheelchair’s responses to everyday obstacles such as ramps, steps, and uneven surfaces.
Real-time execution is achieved using the Beckhoff TwinCAT 3 environment, where control and parameter exchange are enabled through communication based on the TwinCAT ADS protocol. The virtual environment, developed in Unity, provides an immersive and flexible space for generating repeatable terrain profiles and potential interaction scenarios. The wheelchair suspension model, adapted from automotive suspension models (quarter-car, half-car, full-car model), was developed and subjected to multi-stage analysis in the Simulink environment.
The simulator was operationally tested through experiments involving three test subjects who operated the system under identical conditions. To assess fidelity and user-dependent behavior of the system, performance indicators such as lateral deviation from the ideal path, vertical response of the sprung mass, and the user experience of the test subjects were analyzed. The results demonstrate realistic dynamic responses and reveal differences in driving behavior and stability among participants, with a focus on the performance of the simulator.
The final simulator design represents a reliable concept for investigating wheelchair dynamics, user–system interaction, and control strategy development, while also offering potential for future applications in user training and rehabilitation research.
|
