Comparative Study of Optimizer on Single Stage Drone Fault Detection and Identification

Fault Detection and Identification (FDI) is critical to ensuring the safety and reliability of Unmanned Aerial Vehicles (UAVs), particularly as they operate with increasing autonomy in complex environments. While recent advances in deep learning and data-driven approaches have enhanced UAV fault detection capabilities, the role of optimization algorithms within these frameworks remains underexplored. This paper presents a comparative study of three bio-inspired optimization algorithms - Barnacle Mating Optimizer (BMO), Tunicate Swarm Algorithm (TSA), and Rock Hyraxes Swarm Optimization (RHSO) - applied within a unified single stage UAV FDI system. The evaluation is conducted using four performance metrics: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and detection time. Results show that TSA offers the highest fault estimation accuracy, while RHSO provides the fastest detection time. BMO demonstrates a balanced trade-off between accuracy and speed. This study fills a notable gap in UAV FDI research by benchmarking multiple optimizers under consistent conditions on single stage drone FDI. (Abstract)