Flow Field Reconstruction using Optimal Sensor Placement and Deep Learning
Authors: Bhavneet Bali and Mandar Tendolkar
Publishing Date: 22-10-2023
ISBN: 978-81-955020-2-8
Abstract
Fluid flow reconstruction is a crucial task in engineering, environmental science, and fluid dynamics. Accurately predicting and understanding fluid flow patterns is essential for optimizing processes, designing efficient systems, and mitigating potential risks. Traditional methods often rely on complex mathematical models and computationally intensive simulations. However, recent advancements in deep learning and neural networks have shown promising results in tackling this problem. The present work proposes a novel approach to fluid flow reconstruction using neural networks, aiming to develop an efficient and accurate feed-forward neural network model capable of predicting fluid flow behaviour based on available data. An artificial neural network (ANN) is used to capture spatial dependencies in fluid flow data, learning to infer underlying flow dynamics. The model's robustness and generalization capabilities are ensured by carefully curating the dataset and incorporating appropriate data augmentation techniques. The results demonstrate the effectiveness of neural networks in fluid flow reconstruction, with significant improvements in prediction accuracy and efficiency compared to traditional methods. The ability to reconstruct fluid flow patterns accurately from limited or incomplete data has the potential to revolutionize various industries, enabling more informed decision-making, optimizing processes, and improving safety measures. This work contributes to the growing body of knowledge in deep learning for fluid dynamics and offers a promising avenue for further advancements in predicting and understanding fluid flow behaviour.
Keywords
Artificial neural network (ANN), fluid flow reconstruction, Proper orthogonal decomposition (POD).
Cite as
Bhavneet Bali and Mandar Tendolkar, "Flow Field Reconstruction using Optimal Sensor Placement and Deep Learning", In: Satyasai Jagannath Nanda and Rajendra Prasad Yadav (eds), Data Science and Intelligent Computing Techniques, SCRS, India, 2023, pp. 499-508. https://doi.org/10.56155/978-81-955020-2-8-46