Optimal Design of Kinematics Parallel Manipulator Considering Workspace and Control Effort

Abstract

Several studies have reported the design of kinematics parallel mechanisms based on behavioral features; however, the design of this kind of system with six degrees of freedom considering parallelly volumetric behavior together with control effort remains to be accomplished. This work addresses the design of one type of these mechanisms based on two aspects: workspace and control effort. All aspects are considering and optimizing simultaneously through a multiobjective optimization technique based on a bio-inspired algorithm named Elitist Non-Dominated Sorting Differential Evolution Algorithm, which brings about a Pareto front. The workspace is determined using the inverse kinematics constrained boundaries analysis and a mono-objective optimization method. On the other hand, control effort is resolved by calculating the Euclidian norm of each torque signal of the system, which is controlled by a hybrid technique consisting of sliding modes and differential flatness. Finally, relations between the two studied aspects are depicted and analyzed.