Aiming at fast and accurate spatiotemporal prediction of interfacial flow fields, a novel deep learning model combining Convolutional AutoEncoder (CAE) and long short-term memory with spatial and temporal attention (LSTM-STA) is proposed and named as CAE-LSTM-STA (hybrid model of CAE and LSTM-STA) in this article. To enable a fast calculation process, CAE is first utilized to compress the high-dimensional snapshots of flow fields into the low-dimensional latent space by its encoder. Then, the latent space serves as the input feature for LSTM-STA to temporally evolve the state of the low-dimensional latent space. Finally, the low-dimensional latent space at future time steps is fed back into the decoder of CAE to derive the full-order snapshots of the flow fields. The prediction performance of the proposed model is evaluated via two representative benchmark cases, including the dam break case and the rising bubble case. The capability of CAE in dimension reduction is found to be much better than that of the widely used Proper Orthogonal Decomposition, while LSTM-STA is observed that outperforms the original LSTM in multivariant temporal prediction. These promising results indicate that the proposed CAE-LSTM-STA model is able to effectively capture and advance the spatiotemporal characteristics of the interfacial flow fields, therefore making it an advanced surrogate model for fast and precise generation of the temporally continuous interfacial flow fields.
This study performs particle resolved simulations (PRSs) to investigate the behavior of linear shear flow over a near-wall oblate spheroidal particle. Simulations are conducted for particles with aspect ratios (λ) of 1.5, 2, and 4, covering Reynolds numbers (Re) ranging from 1 to 200 and particle inclination angles (θ) from 0° to 180°. The resulting drag, lift, and torque coefficients exhibit unique patterns of variation with respect to θ, showing significant differences compared to particles in unbounded uniform flow. These variations are elucidated by examining the flow field surrounding the particle, alongside the distribution of pressure and skin-friction coefficients on its surface. Based on these findings, correlations for the drag, lift, and torque coefficients are proposed, demonstrating strong agreement with the PRS results across different aspect ratios and Reynolds numbers, thereby providing accurate predictive models for near-wall oblate spheroidal particles in linear shear flow.
This study proposes a novel hybrid intelligent model, POD–CAE–LSTM, to enhance the accuracy and stability of spatiotemporal predictions of unsteady turbulent flows. This model is based on the concept of a triple decomposition of the flow field, in which the flow field is separated into a time-averaged velocity, a coherent component, and a stochastic turbulent component. Proper orthogonal decomposition (POD) is employed to accurately capture the dominant large-scale coherent component of the flow field. Meanwhile, a convolutional autoencoder (CAE) is utilized to achieve efficient nonlinear dimensionality reduction of the stochastic turbulent component. Subsequently, a long short-term memory (LSTM) network with a coupled self-attention mechanism is introduced to perform temporal predictions using the reduced low-dimensional vectors. Finally, the predicted results are reconstructed through the POD modes and the CAE decoder to obtain the future evolution of the flow field. The predictive performance of the proposed model is evaluated using two representative cases: the quasi-periodic flow field of a cylinder wake at a Reynolds number of 1 × 104, and a highly nonlinear single-vortex tornado flow field. The results demonstrate that the proposed model outperforms the widely used POD–LSTM and CAE–LSTM models in terms of both dimensionality-reduction reconstruction accuracy and spatiotemporal prediction capability. These encouraging findings indicate that the proposed POD–CAE–LSTM model successfully integrates the advantages of physical interpretability and deep learning, thereby providing a reliable and efficient solution for the spatiotemporal prediction of complex turbulent flows.
Accurate reconstruction of the dam-break free surface profiles from sparse monitoring water levels is a challenging task in flood disaster warning. To address this issue, this article introduces a novel intelligent model that leverages fully connected autoencoder (FCAE) to acquire low-order latent space representations and random forest (RF) to establish the relationship between sparse monitoring water levels and latent space representations, which is named as FCAE–RF. The proposed model is tested on two benchmark cases, including the advection of a shock profile and the classical dam-break flow. The results indicate that FCAE–RF provides satisfactory prediction performances for both uniform and nonuniform distributions of the monitoring locations. Besides, the adopted FCAE displays better performance than the widely used proper orthogonal decomposition and convolutional autoencoder in the dimension reduction, while the employed RF presents larger superiority than the commonly used ridge regression and multi-layer perceptron in the establishment of the nonlinear relationship. These promising results indicate that FCAE–RF can be a robust model in the application of flood management and control.
30 April 2026 in International Journal of Multiphase Flow