Search

Fault detection and isolation in unmanned aerial vehicle (UAV) propellers are critical for operational safety and efficiency. Most existing fault diagnosis techniques rely basically on traditional statistical-based methods that necessitate better approaches. This study explores the application of untraditional feature extraction methodologies, namely Permutation Entropy (PE), Lempel–Ziv Complexity (LZC), and Teager–Kaiser Energy Operator (TKEO), on the PADRE dataset, which encapsulates various rotor fault configurations. The extracted features were subjected to a Chi-Square (χ2) feature selection process to identify the most significant features for input into a Deep Neural Network. The Taguchi method was utilized to test the performance of the recorded features, correspondingly. Performance metrics, including Accuracy, F1-Score, Precision, and Recall, were employed to evaluate the model’s effectiveness before and after the feature selection. The achieved accuracy has increased by 0.9% when compared with results utilizing traditional statistical methods. Comparative analysis with prior research reveals that the proposed untraditional features surpass traditional methods in diagnosing UAV propeller faults. It resulted in improved performance metrics with Accuracy, F1-Score, Precision, and Recall reaching 99.6%, 99.5%, 99.5%, and 99.5%, respectively. The results suggest promising directions for future research in UAV maintenance and safety protocols. [ABSTRACT FROM AUTHOR]

Unmanned aerial vehicles (UAVs) have gained significant importance due to their wide applicability in modern life. Fault diagnosis plays a crucial role in ensuring their safe and reliable operation. This study evaluated a smart drone's performance under different actuator damage modes. Flight experiments included healthy, minor, moderate, and severe damage with motor base shifting scenarios. Each flight lasted 10 min with a consistent hovering height of 1.2 m. The study underscored the importance of leveraging embedded recorded data to enhance fault diagnosis techniques and improve the reliability and safety of UAV operations. The effectiveness of utilizing the embedded recorded data for fault diagnosis is demonstrated, eliminating the need for conventional test-rigs and reducing associated costs and analysis time. The drone's embedded system recorded flight data, specifically pitching and rolling angles in degrees, which were soft-labeled for analysis. Classification analysis was performed using the ORANGE data-mining program, leveraging the Stacking technique that combined three models: Stochastic Gradient Descent (SGD), K-nearest Neighbor (KNN), and Support Vector Machine (SVM). The individual accuracies of the SGD, KNN, and SVM models were 72.2%, 74.5%, and 71.2%, respectively. However, the stacking technique improved overall accuracy significantly to 96%. The findings highlighted the potential of machine learning techniques and the stacking method in accurately assessing drone performance offering a reliable and cost-effective approach for evaluating UAV performance under varying levels of actuator damage. [ABSTRACT FROM AUTHOR]