Diagnostic of Gravitropism-like Stabilizer of Inspection Drone Using Neural Networks

Abstract

This paper discusses the enhancement of flight stability of using an inspection drone to scan the condition of buildings on low and high altitude. Due to aerial perturbations and wakes, the drone starts to shake and may be damaged. One of the mechanical optimization methods it so add a built-in stabilizing mechanism. However, the performance of this supporting device becomes critical on certain flying heights, thus to avoid losing the drone. The paper is divided in two parts: The description of the gravitropism-like stabilizer and the diagnostic of its status using wavelet transformation and neural network classification.