Deep learning for impact detection in composite plates with sparsely integrated sensors
Source abstract
Abstract
Abstract In this paper, both location and energy of impacts on an anisotropic carbon fiber reinforced plate (CFRP) are detected with the help of deep learning. We introduce sparse low-cost sensor array integration in CFRP plates that allows for structural monitoring of lightweight structures. Using a resin transfer moulding process microelectromechanical systems (MEMS) and piezoelectric transducers (PZT) sensors are integrated into CFRP plates. We developed an automated test bench to perform weight drop impact loadings with impact energies ranging between 0.22–0.56 mJ on a 1 × 1 cm 2 -grid with 441 locations. The obtained sensor signals were processed by means of a short-time fourier transformation and used as input for the training of a deep learning model. This model was implemented with a convolutional neural network. To accelerate the training phase we introduce a coarse analytical model that generates artificial sensor signals we use for pretraining of the neural network. Yielding high prediction accuracies of 99.82% and 98.68% for a correct classification of impact location and energy, respectively, the capability of the proposed approach was demonstrated. Despite their limited resolution the low-cost MEMS accelerometers were able to correctly locate an impact and its energy with 99.76% and 97.04%, respectively. The pretraining led to an increased robustness of the training process. Additionally, for the case of PZT sensors, it also reduced the number of required epochs for convergence significantly.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Structural Health Monitoring Techniques · Ultrasonics and Acoustic Wave Propagation · Smart Materials for Construction
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (49)
Show source-provided references
1. Mixed-mode delamination growth in carbon–fibre composite laminates under cyclic loading10.1016/j.ijsolstr.2004.02.040 ↗
2. Identification of foreign object impact in structures using distributed sensors10.1177/1045389X9400500620 ↗
3. Vibrational based inspection of civil engineering structures
4. An efficient approach for identifying impact force using embedded piezoelectric sensors10.1016/j.ijimpeng.2006.05.004 ↗
5. Identifying impacts in composite plates with piezoelectric strain sensors, part i: theory10.1177/1045389X9800901108 ↗
6. Identifying impacts in composite plates with piezoelectric strain sensors, part ii: Experiment10.1177/1045389X9800901109 ↗
7. Acoustic source localization10.1016/j.ultras.2013.06.009 ↗
8. Improving accuracy of acoustic source localization in anisotropic plates10.1016/j.ultras.2014.03.001 ↗
9. A novel differential time-of-arrival estimation technique for impact localization on carbon fiber laminate sheets10.3390/s17102270 ↗
10. Reconstruction and analysis of impact forces on a steel-beam-reinforced concrete deck10.1007/s11340-016-0188-4 ↗
11. Inverse estimation of impact force on a composite panel using a single piezoelectric sensor10.1177/1045389X16657424 ↗
12. A review on the application of deep learning in system health management10.1016/j.ymssp.2017.11.024 ↗
13. Deep learning and its applications to machine health monitoring10.1016/j.ymssp.2018.05.050 ↗
14. An artificial neural network for low-energy impact monitoring10.1177/089270579400700404 ↗
15. Detection of impacts in composite materials using piezoceramic sensors and neural networks10.1117/12.316918 ↗
16. Structural health monitoring and impact detection using neural networks for damage characterization10.2514/6.2006-2117 ↗
17. An automatic impact monitor for a composite panel employing smart sensor technology10.1088/0964-1726/14/1/027 ↗
18. Impact damage location in composite structures using optimized sensor triangulation procedure10.1088/0964-1726/12/5/017 ↗
19. Determination of impact location on composite stiffened panels10.1088/0964-1726/21/10/105026 ↗
20. Identification of impact force for smart composite stiffened panels10.1088/0964-1726/22/8/085014 ↗
21. A time reversal focusing based impact imaging method and its evaluation on complex composite structures10.1088/0964-1726/20/10/105014 ↗
22. A review of impact damage detection in structures using strain data
23. Convolutional neural network based fault detection for rotating machinery10.1016/j.jsv.2016.05.027 ↗
24. A novel fault diagnosis method for rotating machinery based on a convolutional neural network10.3390/s18051429 ↗
25. Real-time vibration-based structural damage detection using one-dimensional convolutional neural networks10.1016/j.jsv.2016.10.043 ↗
26. Estimation of crowd flow and load on pedestrian bridges using machine learning with sensor fusion10.1016/j.autcon.2020.103092 ↗
27. A convolutional neural network for impact detection and characterization of complex composite structures10.3390/s19224933 ↗
28. Mohammed · 2017 · 33110.1007/978-3-319-50824-5_11 ↗
29. Method for arranging a number of micromechanical acceleration sensors on or in a plastic component and corresponding plastic component
30. Sensor systems for frp lightweight structures: automotive features based on serial sensor products10.3390/s19143088 ↗
31. Communication in the presence of noise10.1109/JRPROC.1949.232969 ↗
32. Mertins · 2010
33. Artificial neural networks: A tutorial10.1109/2.485891 ↗
34. Multilayer feedforward networks with a nonpolynomial activation function can approximate any function10.1016/S0893-6080(05)80131-5 ↗
35. Gradient theory of optimal flight paths10.2514/8.5282 ↗
36. A gradient method for optimizing multi-stage allocation processes
37. LeCun · 1990 · 396
38. Krizhevsky · 2012 · 1097
39. Guo · 2020 · 11710.1007/978-981-15-3863-6_13 ↗
40. Bishop · 200610.1007/978-981-15-3863-6_13 ↗
41. A simple weight decay can improve generalization In
42. Adam: A method for stochastic optimization
43. Krig · 2016 · 31910.1007/978-3-319-33762-3_9 ↗
44. How transferable are features in deep neural networks?
45. TensorFlow: Large-scale machine learning on heterogeneous systems
46. Localization of impact on composite plates based on integrated wavelet transform and hybrid minimization algorithm10.1016/j.compstruct.2017.05.017 ↗
47. Identifying impacts in composite plates with piezoelectric strain sensors, part i: theory10.1177/1045389X9800901108 ↗
48. A two-step impact localization method for composite structures with a parameterized laminate model10.1016/j.compstruct.2018.03.052 ↗
49. Impact localisation in composite plates of different stiffness impactors under simulated environmental and operational conditions10.3390/s19173659 ↗