No abstract

This paper deals with the development of a realtime structural health monitoring system for airframe structures to localize and estimate the magnitude of the loads causing deflections to the critical components, such as wings. To this end, a framework that is based on artificial neural networks is developed where features that are extracted from a depth camera are utilized. The localization of the load is treated as a multinomial logistic classification problem and the load magnitude estimation as a logistic regression problem. The neural networks trained for classification and regression are preceded with an autoencoder, through which maximum informative data at a much smaller scale are extracted from the depth features. The effectiveness of the proposed method is validated by an experimental study performed on a composite unmanned aerial vehicle (UAV) wing subject to concentrated and distributed loads, and the results obtained by the proposed method are superior when compared with a method based on Castigliano’s theorem.

This paper deals with the development of a machine vision based pose estimation system for industrial robots and improving accuracy of the estimated pose using Long Short Term Memory (LSTM) neural networks. To this end, a target object trackable with a monocular camera with ± 90°in all directions was designed and fitted with fiducial markers. The designed placement of these fiducial markers guarantees the detection of at least two non-planar markers thus preventing ambiguities in pose estimation. Moreover, a LSTM network is proposed in order to improve the accuracy obtained from the Levenberg-Marquardt (LM) based pose estimation algorithm during trajectory tracking of the robot's end effector. The proposed method utilizes a LSTM network to extract dynamic features from the pose estimated by the LM algorithm and then feeding it to a regression layer to estimate the correct pose. The effectiveness of the proposed method is validated by an experimental study performed using a KUKA KR240 R2900 ultra robot while following sixteen distinct trajectories based on ISO 9238. The obtained results show that the proposed method significantly improves the pose estimation accuracy and precision of the vision based system during trajectory tracking of industrial robots' end effector.

In this work a monocular machine vision based pose estimation system is developed for industrial robots and the accuracy of the estimated pose is improved via sparse regression. The proposed sparse regression based method is used improve the accuracy obtained from the Levenberg-Marquardt (LM) based pose estimation algorithm during the trajectory tracking of an industrial robot's end effector. The proposed method utilizes a set of basis functions to sparsely identify the nonlinear relationship between the estimated pose and the true pose provided by a laser tracker. Moreover, a camera target was designed and fitted with fiducial markers, and to prevent ambiguities in pose estimation, the markers are placed in such a way to guarantee the detection of at least two distinct non parallel markers from a single camera within ± 90° in all directions of the camera's view. The effectiveness of the proposed method is validated by an experimental study performed using a KUKA KR240 R2900 ultra robot while following sixteen distinct trajectories based on ISO 9238. The obtained results show that the proposed method provides parsimonious models which improve the pose estimation accuracy and precision of the vision based system during trajectory tracking of industrial robots' end effector.

No abstract