A 3D-Printed Dual-Material Compliant Piezoelectric Micromanipulator for Vision-Assisted Local Task-Space Positioning

Abstract

Micromanipulators used in local micro-scale positioning require stable and controllable end-effector motion within a limited workspace. However, compact compliant mechanisms driven by piezoelectric actuators often exhibit nonlinear coupling, hysteresis-related deviation, and systematic task-space errors, especially when motion amplification and multi-material flexible joints are involved. In this work, a dual-material 3D-printed compliant piezoelectric micromanipulator is developed using PLA as the rigid transmission structure and TPU as the compliant flexure joints. The three piezoelectric actuation branches are arranged in a parallel configuration to generate planar motion of a needle-shaped end effector in a local X–Y task space. To improve local positioning accuracy, a vision-assisted task-space compensation strategy is proposed. The method combines quadratic input–output model identification, residual-field feedforward correction, and vision-based endpoint refinement. Multi-point experiments in the local task space show that the proposed strategy progressively reduces end-effector positioning errors compared with model-only control. The results demonstrate the feasibility of the dual-material compliant mechanism and the vision-assisted compensation strategy for local planar needle-tip positioning. Application-level micromanipulation involving target interaction, contact operation, and three-dimensional motion remains future work.