Design Topology Optimization and Kinematics of a Multi-Modal Quadcopter and Quadruped

Abstract

Multi-modal robots exhibit poor performance in locomotion modes due to its additional weight when compared to a single modal robot. Furthermore, the kinematics of a multimodal robot is more complex due to the increased number of actuators hence more prone to errors and malfunctions. The aim of this study was to use topology optimization to reduce the weight of components in a multi-modal robot, optimizing its ability to withstand stresses and loads. An algorithm for the robot's locomotion mode had been implemented using 8-bit microcontrollers. The prototype of this robot had been tested on different types of mock-up terrains. The topology of components of prototype was optimized and 3D printing was used to fabricate them due to their complex and irregular shapes. Utilization of topology optimization reduced the weight of component up to 54.40% while retaining the capability of the component to withstand the stresses and loads, proving the potential of this technique in robot development. The walking algorithm was implemented with trigonometry instead of inverse kinematics due to the limit of 8-bit microcontrollers. Complementary and exponential filters were implemented to improve the performance of flying locomotion, thus featuring auto-levelling in aerial mode. The robot was able to harness the advantage in both flying and walking over speed and energy efficiency after testing in mock-up terrains, making the robot more versatile in many situations.