Date of Award
Open Access Thesis
The robotic arm represents a complicated mechanical system and advanced engineering principles in robotics, enabling precise and efficient manipulation and interaction with objects in a variety of applications, such as manufacturing, healthcare, space exploration, and other industries. In most cases, energy consumption of robotic tools has been given little consideration due to their perceived insignificance compared to the benefits they offer. However, the increasing capacity demands of factories and the expanding use of robotic arms necessitate careful evaluation and reduction of energy consumption. In this regard, given a task configuration, adjusting the arm's movement path emerges as one of the most effective means to modify energy consumption. In this thesis, the energy consumption model of the robotic arm based on kinematics and experimental data is developed. A global path planning approach utilizing genetic algorithm (GA) optimization on a highly parallelized graphics processing unit (GPU) platform is proposed to minimize the energy consumption of a robotic arm by planning its movement path while meeting the task requirements. Real-time data, encompassing arm position, current, and voltage, are acquired through built-in sensors. The Denavit-Hartenberg convention, a method for kinematic analysis of robotic systems, is employed to establish a relationship between torque and energy consumption for each motor of the robotic arm at varied positions. The objective function for the design optimization considers the energy consumption of the robotic arm, path duration, and safety. Numerical experiments for grasping a weight on a table demonstrate that the optimized path for the robotic arm consistently results in the lowest energy consumption when compared to alternative paths designed by human.
Cao, Y.(2023). GPU-Enabled Genetic Algorithm Optimization and Path Planning of Robotic Arm for Minimizing Energy Consumption. (Master's thesis). Retrieved from https://scholarcommons.sc.edu/etd/7474