© The Institution of Engineering and Technology
To date, the authors are not aware of an in-depth investigation about embedded applications of the convolutional neural network (CNN) algorithm on small, lightweight, and low-cost hardware (e.g. microcontroller, FPGA, DSP, and Raspberry Pi) applied to detect faults in structural health monitoring (SHM) systems. In this Letter, the authors implement and evaluate both feasibility and performance of an embedded application of the CNN algorithm on the Raspberry Pi 3. The CNN-embedded algorithm quantifies and classifies dissimilarities between the frames representing healthy and damaged structural conditions. In a case study, the CNN-embedded application was experimentally evaluated using three piezoelectric patches glued onto an aluminium plate. The results reveal an impressively effective 100% hit rate. This performance may significantly impact the design and analysis of CNN-based SHM systems where embedded applications are required for identifying structural damage such as those encountered by aerospace structures, rotating machineries, and wind turbines.
References
-
-
1)
-
7. De Oliveira, M.A., Araujo, N.V.S., Da Silva, R.N., et al: ‘Use of Savitzky-Golay filter for performances improvement of SHM systems based on neural networks and distributed PZT sensors’, Sensors, 2018, 18, (152), pp. 1–18.
-
2)
-
3. Chen, F-C., Jahanshahi, M.: ‘NB-CNN: deep learning-based crack detection using convolutional neural network and naive Bayes data fusion’, Trans. Ind. Electron., 2017, 65, (5), pp. 4392–4400 (doi: 10.1109/TIE.2017.2764844).
-
3)
-
4. Abdeljaber, O., Avci, O., Kiranyaz, M.S., et al: ‘1-D CNNs for structural damage detection: verification on a structural health monitoring benchmark data’, Neurocomputing, 2018, 275, pp. 1308–1317 (doi: 10.1016/j.neucom.2017.09.069).
-
4)
-
1. Liang, C., Sun, F.P., Rogers, C.A.: ‘Coupled electromechanical analysis of adaptive material systems – determination of the actuator power consumption and system energy transfer’, J. Intell. Mater. Syst. Struct., 1994, 5, (1), pp. 12–20 (doi: 10.1177/1045389X9400500102).
-
5)
-
6. Goodfellow, I., Bengio, Y., Courville, A.: ‘Deep learning’ (MIT Press, Boston, USA, 2016, 1st Edn.).
-
6)
-
2. Pan, J., Zi, U., Chen, J., et al: ‘Liftingnet: a novel deep learning network with layerwise feature learning from noisy mechanical data for fault classification’, Trans. Ind. Electron., 2017, 65, (6), pp. 4973–4982 (doi: 10.1109/TIE.2017.2767540).
-
7)
-
5. Balasubramaniyan, C., Manivannan, D.: ‘Iot enabled air quality monitoring system (AQMS) using raspberry Pi’, Indian J. Sci. Technol., 2016, 9, (39), pp. 1–6 (doi: 10.17485/ijst/2016/v9i39/90414).
http://iet.metastore.ingenta.com/content/journals/10.1049/el.2018.0877
Related content
content/journals/10.1049/el.2018.0877
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Correspondence
This article has following corresponding article(s):
interview
Embedded application of convolutional neural networks on Raspberry Pi for SHM