In context
Soft robotics has advanced rapidly, but precise modeling of their large, nonlinear deformations remains a key challenge. Traditional rigid-body models fail for these highly compliant structures, prompting researchers to develop reduced order models that balance accuracy and computational efficiency for practical simulation and control.
What was reported
Researchers at Peking University proposed a reduced order model for a soft robotic surface capable of freeform deformation. The model uses specialized beam elements to discretize the structure: thin-plate beam elements for elastic polyimide films and slender beam elements for liquid crystal elastomer actuators. The nonlinear deformation is solved using the Newton-Raphson iteration method within an updated Lagrangian formulation.
An inverse mapping was derived to compute individual actuator contractions from desired global geometric features, such as local curvature and height. This enables direct control inputs for achieving target surface shapes. The model was validated through numerical simulations of a 3×3 grid soft robotic surface with 42 independently controlled actuators.
Simulations demonstrated accurate prediction of large-deformation behavior, including cylindrical bending, with an average computation time of 13.96 seconds on a standard desktop CPU (i9-13900K) over 100 iteration steps. The model also accounted for passive elongation of actuators due to material soft elasticity.
Why it mattered
This work provides an efficient computational framework for soft robotic surface modeling, simulation, and control. By reducing computational complexity while maintaining accuracy, it facilitates real-time control and simulation training, advancing the practical deployment of soft robots in industrial and interactive applications.
"The reduced order model provides an efficient computational framework for the modeling, simulation, and control of the soft robotic surface."
Source: 《机器人》期刊 (robot.sia.cn) · Published 2025-09-16 · “基于梁单元的曲面软体机器人简化力学模型”