Industrial robots are widely employed in various production and manufacturing processes. As a typical parallel robot, the Delta robot is extensively used in packaging and sorting tasks across industries such as food, pharmaceuticals, and electronics. This widespread application is primarily attributed to its unique architecture, in which the actuators are mounted on a fixed base, enabling high-speed and high-acceleration movements. To achieve high acceleration capabilities while minimizing energy consumption, Delta robots are typically designed with lightweight structures. However, this often leads to significant vibration issues under high acceleration, which severely deteriorate their operational accuracy and limits their potential for application in precision-critical domains. To tackle these challenges, this thesis investigates the problem from three key aspects: parameter optimization, controller design, and trajectory optimization. The aim is to improve the operational accuracy of the Delta robot while minimizing residual vibrations, thereby broadening its applicability in precision-critical applications. More precisely, a parameter optimization approach is first proposed for Delta robots, where the kinematic, rigid-body dynamic, and elastodynamic performances are simultaneously considered. Then, when users have access to the control system of Delta robots, two controllers are designed to achieve trajectory tracking when the robots encounter different obstacles, such as model uncertainties and unavailable velocity information. Furthermore, to mitigate residual vibrations, the rigid-flexible coupling dynamic model of the Delta robot is established, and a vibration suppression controller is designed based on this model. However, since the proposed vibration suppression controller requires real-time vibration signal measurement, it increases the cost due to the need for additional vibration sensors. To address this, an input shaper is designed to mitigate residual vibrations by modifying only the reference trajectories. An iterative learning controller is proposed to achieve high-precision trajectory tracking of the Delta robot. Since controller-based accuracy improvement methods for the Delta robot require access to low-level controllers, which users typically cannot redesign or modify. Additionally, the input shaper inevitably increases traversal time and may lead to trajectory deformation. Therefore, a trajectory auto-generation and optimization approach is proposed to simultaneously ensure vibration suppression and accuracy improvement.
Mingkun Wu (Thu,) studied this question.