A novel class of reconfigurable parallel kinematic manipulators: Concepts and Fourier-based singularity analysis. (November 2020)

• Record Type: Journal Article
• Authors: Camacho-Arreguin, Josue; Wang, Mingfeng; Dong, Xin; Axinte, Dragos
• Is Part Of: Mechanism and machine theory. Volume 153(2020)
• Publication Date: 2020-11
• Subjects: Reconfigurable parallel kinematic manipulators -- Singularity analysis -- Fourier analysis; Machine theory -- Periodicals; Machinery -- Periodicals; Machines -- Périodiques; Génie mécanique -- Périodiques; Machine theory; Machinery; Periodicals; 621.81
• DOI: 10.1016/j.mechmachtheory.2020.103993
• Languages: English
• ISSNs: 0094-114X

Abstract

Highlights: Reconfiguration of upper joints can enhance stiffness and singularity avoidance. Fourier analysis can be used to obtain more information about the Jacobian behaviour. Fourier analysis combined with genetic algorithms can optimise PKMs performance. Abstract: In-situ interventions in complex environments require new concepts of Parallel Kinematic Manipulators (PKMs) that present higher versatility for adapting to unstructured working environments without affecting their advantageous characteristics. To address this opportunity a novel class of adaptive PKMs is proposed, the upper joints are repositionable on the moving platform and the lower joints are free from a base. These characteristics give to the proposed PKMs the capability to modify its workspace and performance (i.e. stiffness and singularity avoidance) to work on non-conventional manufacturing/repair environments. Further, the kinematic and workspace models of the proposed PKMs are introduced. More importantly, a unique method for singularity analysis is introduced, as the versatility to modify the workspace comes with the disadvantage of singularities that are not constant. To address this issue a strategy based on the Fourier transform is introduced, the dominant frequencies are identified and configurations with lower dominant frequencies are weighted to stabilise the Jacobian of the parallel mechanism. Finally, a set of validations are presented for proving the proposed method of singularity analysis by FEA with an error smaller than 1%.