Response Surface Optimization for Compliant Joint of Humanoid Robot Using ANSYS - Design of Experiment

Abstract

Objective: A Compliant Joint of humanoid robot is a spring-loaded assembly, which is used to interact safely with the environment, and it helps to stabilized sudden shock and vibration in the robotic system. At the moment, compliant joints are required to optimize their size and dimensions which result into optimized weight and factor of safety of humanoid robot. Methods: Analysis is carried out using Response Surface Methodology (RSM) and Multi-Objective Genetic Algorithm (MOGA) using ANSYS. The current study employed goal-driven optimisation using ANSYS Workbench to minimise weight and achieve the required factor of safety range for the compliant joint. To find out range of variables such as rim thickness, shaft diameter, base thickness, module thickness and spoke thickness affecting on responses such as factor of safety and geometrical mass of compliant joint single factor single response parametric analysis is carried out. Findings: Based on trend of preliminary analysis variable range and combinations are selected to study interaction effect of parameters to obtain favorable factor of safety and low geometrical mass. The optimized compliant joint is compared with various design and validated through the developed actual module. Novelty: Eventually, the geometry mass of the compliant joint was reduced from 0.8604 kg to 0.6449 kg, resulting in a lighter weight (24.06% reduction) with a 1.7533 factor of safety and more compact in size (outer diameter is shrink from 142 to 126 mm). Keywords: Compliant joint, Goal driven optimization, Response surface optimization, Design of experiment, Humanoid robot