Fractional Order Extended State Observer Enhances the Accuracy of Estimation for Flexible Joint Single-link Robot

作者：Hakam Marwan, A. Humaidi, Huthaifa Al-Khazraji · 发表于：Journal of Robotics and Control (JRC) · 年份：2025 · DOI：10.18196/jrc.v6i3.26318 · 被引用次数：1

This study focuses on analyzing and evaluating the performance of three types of Extended State Observers (ESOs)—Linear ESO (LESO), Nonlinear ESO (NESO), and Fractional-Order ESO (FOESO)—within a flexible joint robotic system. ESOs are a key component of Active Disturbance Rejection Control (ADRC) due to their ability to estimate system states and internal and external disturbances without requiring an accurate system model, making them particularly suitable for nonlinear and complex systems. This research aims to provide a quantitative comparison among the three observer types under parameter uncertainties and external disturbances using MATLAB/SIMULINK simulations. By reducing the RMSE, a Genetic Algorithm (GA) finds the optimal observer parameters. The results indicate that FOESO significantly outperforms the other observers in terms of estimation accuracy, achieving a remarkably low RMSE of 1.06×10⁻¹⁶ for position estimation, compared to 9.16×10⁻¹³ for NESO and 2.17×10⁻¹¹ for LESO. These findings highlight FOESO's superior robustness in managing nonlinearities and disturbances, making it a promising solution for high-precision robotic applications.