Akademik Veri Yönetim Sistemi
8th International Congress on Human-Computer Interaction, Optimization and Robotic Applications, ICHORA 2026, Ankara, Türkiye, 21 - 23 Mayıs 2026, (Tam Metin Bildiri)
The deployment of mobile robots in industrial environments characterized by limited space necessitates the implementation of energy optimization strategies to ensure continuous operation. Mecanum wheels are frequently utilized due to their high maneuverability; however, their unique 45° roller structure creates natural internal friction and slippage, resulting in higher energy consumption than standard wheels. This study investigates the parametric optimization of a mecanum-wheeled mobile robot navigating a constrained trajectory defined by strategic obstacles. By employing the Taguchi method to minimize the number of experiments, the parameters of linear velocity, angular velocity, and payload capacity were systematically modified to analyze their impact on total energy consumption. Experimental data was utilized to train machine learning models, implicitly capturing the nonlinear friction losses that are challenging to model analytically. Among the evaluated algorithms, Random Forest yielded the highest prediction accuracy with an $\mathbf{R}$-squared score of 0.903 and a MAPE of 44.97. A simulation environment mirroring the real-world constrained path demonstrated that adaptively adjusting velocity parameters minimize energy consumption successfully. Theoretical energy consumption values aligned closely with predictions, validating the proposed data-driven framework. The findings provide a practical optimization methodology for enhancing the energy efficiency of mecanumwheeled systems under varying payload and velocity conditions in real-world industrial applications.