Bird Mating Optimizer for Modeling of Flexible Manipulator System

This paper introduces a methodology for modeling a flexible manipulator using the System Identification technique via Bird Mating Optimizer (BMO). The interest in studying flexible manipulators has grown significantly owing to their advantages, such as lightweight design and rapid system response. However, these manipulators exhibit vibrations due to their low stiffness when subjected to disturbances. Moreover, as their speed increases during maneuvers, these unwanted vibrations become more pronounced. Hence, accurately modeling and controlling the nonlinear dynamics of the system is of utmost importance. The primary objective of this study is to develop a precise dynamic model and employ an intelligent optimization technique to address the challenges associated with flexible manipulators. Experimental input-output data for endpoint acceleration were gathered from prior research. To build the dynamic system model, the System Identification technique with the AutoRegressive with eXogenous (ARX) model structure was utilized. In this study, Bird Mating Optimizer (BMO) was introduced specifically for modeling flexible manipulators. Subsequently, the performance and effectiveness of BMO was evaluated and compared against the Particle Swarm Optimization (PSO) algorithm. The results obtained demonstrate that BMO outperforms PSO, achieving the smallest mean square error (MSE) of 4.19 x 10
^-7
.