Sensor Integration and Data Exploitation of Structural Health Monitoring Network Integrated on a Unmanned Aerial Vehicle (UAV)

Abstract

Unmanned Aerial Vehicles (UAVs) are flying robots that require little or no human control while flying. The number of UAV has growth exponentially due to rise in applications such as video taking, surveillance and delivering cargo to customers. Despite the promise of the technology, UAV are not fully reliable currently, either civil or military. They lack of sensing mechanisms to detect issues in operation, such as impacts, hard landings and overload, together with the difficulties to detect structural problems during the UAV operation could compromise not only the safety of the UAV, also all along the entire flying area. This paper presents the design, development implementation, and validation of a Structural Health Monitoring (SHM) system applied to the rear fuselage of the UAV developed by INTA for R&D activities (MILANO). The rear fuselage, a 2.5 meters carbon/epoxy structure with frames and stringers, was instrumented with fiber optic sensors and PZT. Two different types of fiber optic sensors (FOS) were considered: Fiber Bragg Gratings (FBG) and distributed fiber sensing based on Rayleigh backscatter. The objective of the fiber optic sensor network is to detect the changes induced on the strain field due to damage appearance. Different algorithms, such PCA and Artificial Neural Network (ANN) will be discussed. Validation will be performed with measurements on a real structure with increasing damages by removing rivets progressively at different load levels. Long Short-Term Memory (LSTM) networks, a type of Recurrent Neural Networks (RNN), will be proposed for accurate damage detection with the FBG sensing network in real time operations.