Modeling, Simulation and Optimization of the DLR Scout Rover to Enable Extraterrestrial Cave Exploration

Kurzfassung

In-situ exploration of extraterrestrial area of scientific interest, such as caves on the Moon or Mars, remains largely impossible with the state of the art wheeled rovers. One possibility to combine advantages of wheeled and legged robots is the rimless wheel. DLR's Institute of System Dynamics and Control is currently developing a modular rover with rimless wheels, called Scout. It is aimed to be capable to explore caves and other areas that are difficult to access. Development and analysis of robotic concepts at the state of the art sets high requirements on modeling and simulation, among other reasons because testing in relevant environment is only possible to a certain extent. The current trend in robotics, for space applications but also in general, aims for faster and cheaper development where digital simulation models are preferred over hardware prototypes. They're versatile and can be used as basis of optimization campaigns. Parameters and subsystems can be optimized before assembling the first prototypes. Modeling and simulation also permits to identify strengths and weaknesses early in the project. In later phases of the project and during operation, the simulations are valuable for analysis of engineering data. The modeling needs to be detailed and accurate for all this to be possible. It constantly accompanies the whole system and takes the same development steps. This work is going to detail the effort in modeling and simulation of the Scout rover as a multibody system. The whole process follows the V-shaped development process. This starts with the detailed modeling, analysis and optimization of the rimless wheels and other parts that are central to the Scout rover. Places where simplifications are valid, are shown as well. It continues with simulation tests on the full system along with optimization of parameters and subsystems. A first prototype of the Scout rover has been built 1.5 years after the start of the project. It is already exceeding the requirements and expectations. Such good results would not have been possible this fast without simulation and optimization. Refinement of the simulation model and validation against the hardware are the next steps which are currently on-going.