In the realm of scientific research and exploration of the unknown, space exploration stands out as one of the most captivating and crucial fields of our time. Space exploration involves a diverse range of spacecraft designed for various missions and objectives. Among these vehicles, rovers are automated machines specifically designed to traverse the surfaces of planets and moons. Notable examples of rovers include the Curiosity rover on Mars and the Perseverance rover.
However, rovers have inherent limitations, such as slow movement speed, exposure to hostile environments that can reduce their lifespan, limited capacity for carrying scientific instruments, and challenging terrains that can hinder their mobility.
In the face of these challenges, spherical robots emerge as a promising alternative for space exploration. Their integration into the aerospace domain has become a significant focus over the past decade. They stand out for their ability to combine speed, collision resistance, and efficiency while requiring a minimal number of actuators.
There are several types of spherical robots, each distinguished by its method of actuation. In this study, we focus on barycentric spherical robots (BSRs), which move by adjusting their center of mass, with particular emphasis on ARIES, a 2-degree-of-freedom spherical robot specially designed for lunar cave exploration.
The goal of this research lies in modeling ARIES using the Lagrange method, adopting three distinct approaches : a simplified approach assuming independent transverse and longitudinal movements, a complete approach without any simplifications, and an approach aimed at reducing the system’s dynamics. Concurrently, we explore the control of ARIES using sliding mode control, while validating it through comparison with computed torque control. We also compare ARIES’ mechanism to that of a double pendulum, a more common concept in the literature.
The results reveal that sliding mode control offers remarkable performance for systems like ARIES. Additionally, they highlight that the complete dynamics yield superior performance, even though decoupled dynamics reduce the number of parameters and, consequently, computation time.
This study makes a significant contribution to our understanding of spherical robots, spanning from their modeling to their control. These findings are poised to guide the future development of spherical robots and inform design and control choices across a wide range of applications.
| Date | 16 Nov 2023 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | David St-Onge (Supervisor) & Maarouf Saad (Co-supervisor) |
|---|
Diouf, A. N. N. (Author),
St-Onge (Supervisor) &
Saad (Co-supervisor),
16 Nov 2023Student thesis: Master's thesis › Master in Engineering: Engineering