返回
Physical and Data-Driven Optimization for Thin-Walled Manipulator Based on Motion Reliability
DOI:10.1109/TMECH.2025.3553522.png)
摘要
En 中文
In automated mechanical systems, there are strict requirements for the efficiency and accuracy of the movement of certain specific mechanisms, which are often achieved through feedback control. To determine how to select control parameters for a flexible thin-walled manipulator to achieve high-quality movement, we conducted optimization research based on motion accuracy reliability. We have constructed a coupled system dynamics model that integrates the mechanism and motor, allowing the introduction of a controller during the simulation process and considering the randomness of the load. Taking into account that the scale of solving flexible dynamics will result in huge computational costs for subsequent reliability and optimization analysis, we have introduced a data-driven numerical integration scheme. This scheme significantly reduces the time required for dynamics simulation while ensuring acceptable accuracy and adaptability. In the case study, the simulation time can be reduced by about 20 times through the support of data-driven technology while accurately capturing the motion patterns of the manipulator, including its flexible vibrations. Further optimization results indicate that we can improve motion efficiency by approximately $10 \%$ by designing control strategy parameters, while ensuring motion accuracy reliability and stability.
Keyword:
Deep neural network
electromechanical dynamics
motion accuracy reliability
thin-walled manipulator
期刊
I
IF:
0
论文数:
112
被引数:
0
机构
引用论文
A novel variable restitution coefficient model for sphere–substrate elastoplastic contact/impact process一种用于球-基体弹塑性接触/冲击过程的变量恢复系数模型
A multibody-dynamics based method for the estimation of wear evolution in the revolute joints of mechanisms that considers link flexibility一种基于多体动力学的方法,用于估计考虑连杆柔性的机构的旋转关节中的磨损演化
Flexible multibody dynamic analysis of shells with an edge center-based strain smoothing MITC method

