A robust optimization framework for design of robotic system with kinematic and dynamic criteria

Industrial robot, as one class of digitalized intelligent equipment, plays a significant role in enhancing production efficiency and quality through implementing desired kinematic precision and reliable performance for modern high-tech industries. This study proposes a robust optimization framework to account for the kinematic and dynamic uncertainties in industrial robotic systems. The design objective is established with the kinematic extremes and means of motion by incorporating the statistical moments of positioning accuracy and torque error in the joints. The nondeterministic kinematic and dynamic optimization is carried out by integrating the moment-based method and computational optimization techniques. Specifically, the mixed-degree cubature formula is adopted to evaluate the objective function under given design parameters. The optimization algorithm is utilized to achieve a best possible design. The moment sensitivity of motion error with respect to the uncertain parameters is derived by numerical integration. In this study, three practical design examples are provided to demonstrate the effectiveness of the proposed kinematic and dynamic robust optimization methods. The computational and experimental results indicate that the proposed robust optimization framework enables to reduce motion error and lower its sensitivity to uncertainties, thereby achieving reliable improvements in both motion accuracy and system robustness.