A Novel Parameter Estimation Method for Pneumatic Soft Hand Control Applying Logarithmic Decrement for Pseudo-Rigid Body Modeling

Abstract

Controlling soft robots, especially soft hand grasping, is complex due to their ubiquitous deformation, prompting the use of reduced model-based controllers to provide sufficient state information for high dynamic response control performance. However, most modeling techniques face computational efficiency and complexity of parameter identification issues. To alleviate this, a paradigm coupling an analytical modeling approach based on pseudo-rigid body modeling and the logarithmic decrement method (PRBM + LDM) for parameter estimation is proposed. Using a soft robot hand test bed, the PRBM + LDM model for a closed-loop position controller is applied and is compared with a simple proportional–integral–derivative controller (PID controller) static shape control of soft continuum robots using deep visual inverse kinematic models. Furthermore, the PRBM + LDM model-based force controller is compared with simple constant pressure grasping control by pinching tasks on low-weight, small objects—a screwdriver, a potato chip, and a brass coin. The PRBM + LDM-based position controller outperforms the simple PID position controller, and the PRBM + LDM-based force controller achieves a higher success rate than the constant pressure grasping control in the pinching tasks. In conclusion, the PRBM + LDM modeling technique proves to be a convenient and efficient way to model the dynamic behavior of soft actuators closely and can be applied to build high-precision position and force controllers.