Data-driven Controller Selection for Trajectory Control of Unmanned Ground Vehicles Through an Integrated Fuzzy Decision Framework

Technological advances have made unmanned ground vehicles (UGVs) an integral part of modern life. Interest in UGVs has grown significantly, particularly in applications where human safety is a priority or where operations must be performed in environments inaccessible to humans. In recent years, research on UGVs has focused on key topics such as obstacle avoidance, path planning, mapping, localization, navigation, and trajectory tracking. Among these, trajectory tracking remains one of the most challenging problems and has attracted considerable attention from researchers. Furthermore, selecting an appropriate controller for trajectory tracking represents a critical decision-making problem. This study makes an important contribution to the literature, as comprehensive analyses employing multi-criteria decision-making (MCDM) methods for controller selection in UGV trajectory tracking remain scarce. In particular, systematic and comparative evaluations designed to support the controller selection process are still limited. In this context, this study is among the first to simultaneously apply fuzzy AHP and fuzzy ARTASI, two MCDM methods, to address the controller selection problem in UGV trajectory tracking. First, fuzzy AHP was employed to determine the weights of the evaluation criteria. Subsequently, six controllers were evaluated and ranked using the fuzzy ARTASI method based on seven evaluation criteria and linguistic assessments. Finally, the consistency and robustness of the obtained results were verified through a comparative analysis using the fuzzy MABAC, fuzzy TOPSIS, fuzzy VIKOR, and fuzzy MOORA methods. The findings indicate that a hybrid adaptive–sliding mode controller is the most suitable alternative for UGV trajectory tracking.