Design Optimization of a Mars Rover’s Rocker-Bogie Mechanism using Genetic Algorithms

In this work we present our findings in the optimization of a particular suspension system known as the ”rocker-bogie” for our Mars Rover. This type of mechanism has been used on most of the rovers on Mars and has proved to be a simple and elegant design. This work defines the most important performance metrics for a planetary rover. A Genetic Algorithm is implemented and used to optimize the geometry and kinematics of the rover’s wheel suspension system subject to the defined performance metrics. A Genetic Algorithm is chosen for optimization due to its simplicity of implementation and fast convergence to an optimum within a complex and unknown search space. This work shows the effectiveness of the optimization of a rocker-bogie suspension system using a Genetic Algorithm. It also reveals that the resulting system meets all constraints and that significantly reduces the error of individual performance metrics and the overall system. It is shown that the overall fitness of the rover suspension system can be increased by an average of 28% after 100 iterations compared to an initial guess. All performance metrics defined are improved significantly throughout the optimization. The method can be applied to different types of rovers in order to optimize the wheel suspension mechanism’s geometry.