With the wide applications of unmanned aerial vehicle (UAV), operating safety becomes a critical issue. Thus, fault detection (FD) has been focused, which can realize alarm and schedule maintenance in time. Since accurate physical model UAV is usually difficult to obtain flight data with random noise both spatial temporal correlation, huge challenge posed FD. In this article, data-driven multivariate regression approach based on long short-term memory residual filtering (LSTM-RF) proposed...

With the wide applications of unmanned aerial vehicle (UAV) in civilian and military fields, its operational safety has drawn much attention. A series fault detection methods are studied to avoid disasters. Due capabilities strong feature extraction massive flight data processing, deep learning-based have received extensive However, restricted by UAV airborne size, weight, power consumption, a significant challenge is posed deploy these complicated application, which requires run real time....

Unmanned Aerial Vehicle (UAV) can accomplish various specific tasks and play an increasingly essential role in military, industrial civil fields. However, the safety of UAV is lower than that manned aircraft, great economic loss caused due to its relatively high failure rate. Therefore, it significance study anomaly detection method for system. In recent years, deep learning has been widely applied fields outstanding advantages such as strong ability approximate complex functions automatic...

The actuator is critical for controlling the unmanned aerial vehicle (UAV) attitude. With development of sensing technology, convenience collecting high-quality data by sensors has provided favorable conditions data-driven-based UAV fault detection methods. However, there are less historical on a new type UAV, making it challenging to accurately obtain data-driven models. domain distribution (target domain) different from that existing UAVs (source domains), and basic methods difficult...

On-line anomaly detection is critical for the safety of unmanned aerial vehicles (UAVs). However, flight status assessment still depends on ground control stations, which cannot meet time requirement autonomous and safe flight. The lack on-board intelligent systems makes it rather difficult on-line estimation assessment. In order to achieve real-time monitoring UAV enhance reliability UAVs, an embedded system designed address challenging issues in this paper. During flight, sensors key...

Unmanned aerial vehicles (UAVs) are widely used in military and civilian applications. The safety of UAV has been paid more attention. Prognostic Health Management (PHM) can realize the fault prediction during flight make an appropriate response according to potential faults. Therefore, reliability improved. However, existing PHM research on UAVs following two challenges: a) number samples historical data is small, it impossible cover multiple modes meet demand modeling verification, b)...

Satellite telemetry data is regularly received by the ground station, and then staff determines potential operating risk checking real-time data. Among these analysis tasks, anomaly detection necessary to determine failure or early fault. Nowadays, operator usually used historical experience set a fixed alarm threshold judge However, lots of complex abnormal are difficult capture in time accurately setting experiential threshold. Thus, this work, series method proposed based on...

There are almost no on-board intelligent anomaly detection systems in most of the existing unmanned Aerial Vehicles (UAVs), and flight status assessment still depends on ground control station. While, this method can't meet requirement real-time for UAV autonomous safe flight. In order to achieve monitoring status, improve reliability safety UAVs. paper, an on-line non-invasive embedded system (EADS) is designed solve problem complex environment. During flight, whether sensors hardware...

With the widespread application of Unmanned Aerial Vehicle (UAV), more and attention has been paid to analysis flight data, which can realize condition monitoring help improve operational safety. The UAV phase is dynamically switched, identification vital basis for accurate data analysis. However, do not always contain a parameter that be used direct division phases, it difficult give uniform thresholds identification. To automatic phases based on airborne sensor an unsupervised Gaussian...

Unmanned Aerial Vehicles (UAVs) have been widely used in military and civilian applications. Meanwhile, anomaly detection as an essential part of UAV condition monitoring has become particularly critical for maintenance scheduling mission re-planning advance, especially autonomous UAVs. Due to the issue multiple flight modes dynamic switching actual operation, adaptability methods is always challenging when dealing with different trajectories. In this work, a data-driven method data enhanced...

The turbofan engine is the main component of a power system in commercial aircraft, which provides thrust for safe flight aircraft. digital twin builds virtual model corresponding to real by using operating mechanism, historical data, and real-time monitoring data. It can calculate multi-scale high-fidelity data visually, also manage life cycle. One key functions test analysis performance monitoring. This paper realizes function fusing turbofan's mechanism data-driven model. verification...

Monitoring the flight status of aircrafts is crucial for ensuring safe and reliable flights. A global monitoring model a commonly used method to adapt whole process. However, does not accurately capture features when distribution data dynamically transformed with phases, which affects performance model. Therefore, it essential subdivide into different phases by analyzing distributions before building In view this, this paper proposes Divide-and-conquer Gaussian Mixture Model (DcGMM)...

Unmanned Aerial Vehicle (UAV) is widely used in daily life, commercial and military applications, because it flexible cheap. But the news of UAV accidents calls people to focus on Prognostics Health Management (PHM) for UAV, which can predict detect faults during 's flight through telemetry data. PHM also help adjust itself avoid crashes. However, existing models have some common problems. The most outstanding one that they are all suffering from lacking fault amount data difficult support...

Unmanned aerial vehicles (UAVs) have been widely used. To avoid the huge losses caused by high accident rate of UAVs, it is significant to carry out research on UAV condition monitoring (CM) find anomalies in UAVs time. Establishing a baseline model normal operation status from flight data provides possibility realize CM. However, uncertainty multiple flights makes difficult enable generalization ability model. Therefore, this paper proposes modeling method with representation based for...

With the increasing use of Unmanned Aerial Vehicles (UAVs) in various fields, coordinated execution tasks by multiple UAVs has become an important development trend future. To avoid collision with each other during flight and ensure safety, it is essential to be able achieve high-precision, real-time airborne UAV object detection. In this work, a detection method called Mob-YOLO proposed. Based on high-performance model YOLOv4, MobileNetv2, lightweight convolutional neural network, used...

Due to the excellent performance and cost-effective, unmanned aerial vehicle (UAV) has been widely used in civil military fields. But accident rate of UAV is much higher than that manned aircraft. Therefore, sensor data monitoring become a research hotspot, which can further support Prognostics Health Management (PHM). However, on-board computing resources power are limited, most state-of-the-art methods only be operated on ground. A huge challenge presented real-time condition monitoring....

With the integration of Artificial Intelligence (AI) and robots, unmanned autonomous systems (UAS) have been widely used both in civil military fields. However, complexity UAS increases with improvement its ability, which puts forward higher requirements for online condition monitoring (OCM) to timely confirm safety availability UAS. To achieve OCM, an embedded OCM instrument combing perception, intelligent algorithms, edge computing should be developed, where perception is responsible...