Fusion of Multiple Data Sources for Vehicle Crashworthiness Prediction Using CycleGAN and Temporal Convolutional Networks

Computer-aided engineering (CAE) models play a pivotal role in predicting crashworthiness of vehicle designs. While CAE models continue to advance in fidelity and accuracy, an inherent discrepancy between CAE model predictions and the responses of physical tests remains inevitable, due to assumptions or simplifications made in physics-based CAE models. Machine learning (ML) models have shown promising potential in improving the prediction accuracy of CAE models. Nevertheless, the scarcity of vehicle crash data poses a significant challenge to the training of such ML models. This paper aims to overcome these challenges by fusing multiple data sources from two different types of vehicles. More specifically, the cycle-consistent generative adversarial neural networks (CycleGAN) are first employed to translate features of time-series test data from one domain (the first vehicle type) to another (the second vehicle type) using cycle consistency loss. Such a translation allows for the generation of synthetic crash test data for the second vehicle type by leveraging existing tests from both the first and second vehicle types. In parallel, an initial temporal convolutional network (TCN) model is trained using CAE simulation data and physical test data of the first vehicle type. This pre-trained TCN model is then fine-tuned using three sources of data from the second vehicle type, namely the CAE data, test data, and the augmented virtual test data generated using CycleGAN. Through this data fusion, the crashworthiness prediction accuracy of the second vehicle type can be improved. The essence of the proposed method involves domain translation across two different yet potentially interrelated vehicle types. This is accomplished by leveraging insights gained from the first vehicle type through transfer learning, coupled with data augmentation techniques. The proposed method is demonstrated by a real-world case study with a small-size SUV and a medium-size SUV. Results show substantial enhancement in the predictive performance of the medium-size SUV model.