Sequential Modelling in Data-Driven Approach for Structural Health Monitoring by Recurrent Convolutional Neural Networks

---

Constituent Research Paper Publications

I] On Stability of Spatially Convoluted Sequential Modelling for Data-Driven Approach in Structural Health Monitoring

Conference: 20th World Conference on Non-Destructive Testing
Location &/Or Date: South Korea, Seoul - Feb. 2022
Co-authors: Ewald V., Goby X., Groves R.M., Benedictus R.
Laboratory: TU Delft Aerospace NDT Lab
Status: Working Paper Personal Role/Contribution: Development of distinct Recurrent CNN architecture spatio-temporal models. Evaluation of the performance & level of instability of each architecture @ time of inference.

&

II] On Stability of Sequential Modelling in Data Driven Predictive Maintenance: A Study Case Using CNN-LSTM for DeepSHM

Journal: Intl Journal of Structural Health Monitoring Location &/Or Date: Dec. 2021
Co-authors: Ewald V., Goby X., Groves R.M., Benedictus R.
Laboratory: TU Delft Aerospace NDT Lab
Status: In Press Personal Role/Contribution: Development & fine-tuning of CNN-LSTM for spatio-temporal forecasting on time-frequency representation data in a continuous learning manner.