Global stiffness modeling and optimization of a 5-DOF parallel mechanism

摘要

Global Stiffness design and optimization of parallel mechanisms can be a difficult and time-consuming exercise in parallel robot design, especially when the variables are multifarious and the objective functions are too complex. To address this issue, optimization techniques based on kinetostatic model and genetic algorithms are investigated as the effective criteria. First, a 5-DOF parallel mechanism with a passive constraining leg and five identical legs with prismatic actuators for machine tool is proposed, and its corresponding inverse kinematics, Jacobian matrices and global velocity equation are derived. Second, with the kinetostatic model, the mean value and the standard deviation of the trace of the global compliance distribution are proposed as these two kinetostatic performance indices. Finally, the effectiveness of this optimization design methodology for global stiffness indices is validated with simulation.