Mechanical Joint Compliance Estimation of a Robotic Arm using Applied System Identification Methods

In this study, a mathematical model of a single-degree-of-freedom robotic arm is estimated using system identification method with numerical input and output data obtained from Simscape Multibody. The mechanical model of the robotic arm is developed using the Simscape Multibody modeling environment, which facilitates the construction and simulation of complex multibody mechanical systems. The focus is on leveraging the simulation capabilities of Simscape Multibody to obtain accurate input and output signals that reflect the system's dynamic behavior. These signals are then employed for the purpose of estimating mathematical models that represent the underlying dynamics of the robotic arm. The system identification process is performed using the Transfer Function Estimation (TFEST) method, which is a continuous-time identification technique. Mathematical models of the robot arm are estimated for the angular displacement, angular velocity, angular acceleration signals collected from the joint of the robot arm. The comparative results demonstrate that the TFEST method accurately captures the dynamic characteristics of the robotic arm, aligning closely with the behavior observed in both the analytical and Simscape Multibody simulation models. The findings of this study highlight the applicability of the TFEST method as a reliable tool for modeling the dynamic behavior of robotic systems.