-
Application of probabilistic modeling and automated machine learning framework for high-dimensional stress field
Authors:
Lele Luan,
Nesar Ramachandra,
Sandipp Krishnan Ravi,
Anindya Bhaduri,
Piyush Pandita,
Prasanna Balaprakash,
Mihai Anitescu,
Changjie Sun,
Li** Wang
Abstract:
Modern computational methods, involving highly sophisticated mathematical formulations, enable several tasks like modeling complex physical phenomenon, predicting key properties and design optimization. The higher fidelity in these computer models makes it computationally intensive to query them hundreds of times for optimization and one usually relies on a simplified model albeit at the cost of l…
▽ More
Modern computational methods, involving highly sophisticated mathematical formulations, enable several tasks like modeling complex physical phenomenon, predicting key properties and design optimization. The higher fidelity in these computer models makes it computationally intensive to query them hundreds of times for optimization and one usually relies on a simplified model albeit at the cost of losing predictive accuracy and precision. Towards this, data-driven surrogate modeling methods have shown a lot of promise in emulating the behavior of the expensive computer models. However, a major bottleneck in such methods is the inability to deal with high input dimensionality and the need for relatively large datasets. With such problems, the input and output quantity of interest are tensors of high dimensionality. Commonly used surrogate modeling methods for such problems, suffer from requirements like high number of computational evaluations that precludes one from performing other numerical tasks like uncertainty quantification and statistical analysis. In this work, we propose an end-to-end approach that maps a high-dimensional image like input to an output of high dimensionality or its key statistics. Our approach uses two main framework that perform three steps: a) reduce the input and output from a high-dimensional space to a reduced or low-dimensional space, b) model the input-output relationship in the low-dimensional space, and c) enable the incorporation of domain-specific physical constraints as masks. In order to accomplish the task of reducing input dimensionality we leverage principal component analysis, that is coupled with two surrogate modeling methods namely: a) Bayesian hybrid modeling, and b) DeepHyper's deep neural networks. We demonstrate the applicability of the approach on a problem of a linear elastic stress field data.
△ Less
Submitted 11 April, 2023; v1 submitted 15 March, 2023;
originally announced March 2023.
-
Stress field prediction in fiber-reinforced composite materials using a deep learning approach
Authors:
Anindya Bhaduri,
Ashwini Gupta,
Lori Graham-Brady
Abstract:
Computational stress analysis is an important step in the design of material systems. Finite element method (FEM) is a standard approach of performing stress analysis of complex material systems. A way to accelerate stress analysis is to replace FEM with a data-driven machine learning based stress analysis approach. In this study, we consider a fiber-reinforced matrix composite material system and…
▽ More
Computational stress analysis is an important step in the design of material systems. Finite element method (FEM) is a standard approach of performing stress analysis of complex material systems. A way to accelerate stress analysis is to replace FEM with a data-driven machine learning based stress analysis approach. In this study, we consider a fiber-reinforced matrix composite material system and we use deep learning tools to find an alternative to the FEM approach for stress field prediction. We first try to predict stress field maps for composite material systems of fixed number of fibers with varying spatial configurations. Specifically, we try to find a map** between the spatial arrangement of the fibers in the composite material and the corresponding von Mises stress field. This is achieved by using a convolutional neural network (CNN), specifically a U-Net architecture, using true stress maps of systems with same number of fibers as training data. U-Net is a encoder-decoder network which in this study takes in the composite material image as an input and outputs the stress field image which is of the same size as the input image. We perform a robustness analysis by taking different initializations of the training samples to find the sensitivity of the prediction accuracy to the small number of training samples. When the number of fibers in the composite material system is increased for the same volume fraction, a finer finite element mesh discretization is required to represent the geometry accurately. This leads to an increase in the computational cost. Thus, the secondary goal here is to predict the stress field for systems with larger number of fibers with varying spatial configurations using information from the true stress maps of relatively cheaper systems of smaller fiber number.
△ Less
Submitted 31 October, 2021;
originally announced November 2021.
-
Language Independent Speech Emotion and Non-invasive Early Detection of Neurocognitive Disorder
Authors:
Susmita Bhaduri,
Anirban Bhaduri,
Rajib Sarkar
Abstract:
Emotions(like fear,anger,sadness,happiness etc.) are the fundamental features of human behavior and governs his/her mental health. The subtlety of emotional fluctuations can be examined through perturbation in conversations or speech. Analysis of emotional state of a person from acoustical features of speech signal leads to discovery of vital cues determining his or her mental health. Hence, it's…
▽ More
Emotions(like fear,anger,sadness,happiness etc.) are the fundamental features of human behavior and governs his/her mental health. The subtlety of emotional fluctuations can be examined through perturbation in conversations or speech. Analysis of emotional state of a person from acoustical features of speech signal leads to discovery of vital cues determining his or her mental health. Hence, it's an important field of research in the area of Human Computer Interaction(HCI). In a recent work we have shown that how the contrast in Hurst-Exponent calculated from the non-stationary and nonlinear aspects of "angry" and "sad" speech(spoken in English language) recordings in the Toronto-Emotional-Speech-Set(TESS) can be used for early detection and diagnosis of Alzheimer's Disease. In this work we have extended the work and extracted Hurst-exponent for the speech-signals of similar emotions but spoken in German language. It has been observed that the Hurst-exponent efficiently segregates the contrasting emotions of "anger" and "sadness" in the speech spoken in German language, in similar fashion it has been doing for English speech. Hence it can be concluded that the Hurst-exponent can differentiate among speech spoken out of different emotions in language-independent manner. We propose algorithm for a language-independent application for early non-invasive detection of various severe neurocognitive-disorders like Alzheimer's Disease, MND(motor-neuron-disorder), ASD(autism-spectrum-disorder), depression, suicidal-tendency etc. which is not possible with the state of the art medical science.
△ Less
Submitted 3 June, 2021;
originally announced June 2021.
-
An efficient optimization based microstructure reconstruction approach with multiple loss functions
Authors:
Anindya Bhaduri,
Ashwini Gupta,
Audrey Olivier,
Lori Graham-Brady
Abstract:
Stochastic microstructure reconstruction involves digital generation of microstructures that match key statistics and characteristics of a (set of) target microstructure(s). This process enables computational analyses on ensembles of microstructures without having to perform exhaustive and costly experimental characterizations. Statistical functions-based and deep learning-based methods are among…
▽ More
Stochastic microstructure reconstruction involves digital generation of microstructures that match key statistics and characteristics of a (set of) target microstructure(s). This process enables computational analyses on ensembles of microstructures without having to perform exhaustive and costly experimental characterizations. Statistical functions-based and deep learning-based methods are among the stochastic microstructure reconstruction approaches applicable to a wide range of material systems. In this paper, we integrate statistical descriptors as well as feature maps from a pre-trained deep neural network into an overall loss function for an optimization based reconstruction procedure. This helps us to achieve significant computational efficiency in reconstructing microstructures that retain the critically important physical properties of the target microstructure. A numerical example for the microstructure reconstruction of bi-phase random porous ceramic material demonstrates the efficiency of the proposed methodology. We further perform a detailed finite element analysis (FEA) of the reconstructed microstructures to calculate effective elastic modulus, effective thermal conductivity and effective hydraulic conductivity, in order to analyse the algorithm's capacity to capture the variability of these material properties with respect to those of the target microstructure. This method provides an economic, efficient and easy-to-use approach for reconstructing random multiphase materials in 2D which has the potential to be extended to 3D structures.
△ Less
Submitted 3 February, 2021;
originally announced February 2021.
-
Detecting tala Computationally in Polyphonic Context - A Novel Approach
Authors:
Susmita Bhaduri,
Anirban Bhaduri,
Dipak Ghosh
Abstract:
In North-Indian-Music-System(NIMS),tabla is mostly used as percussive accompaniment for vocal-music in polyphonic-compositions. The human auditory system uses perceptual grou** of musical-elements and easily filters the tabla component, thereby decoding prominent rhythmic features like tala, tempo from a polyphonic composition. For Western music, lots of work have been reported for automated dru…
▽ More
In North-Indian-Music-System(NIMS),tabla is mostly used as percussive accompaniment for vocal-music in polyphonic-compositions. The human auditory system uses perceptual grou** of musical-elements and easily filters the tabla component, thereby decoding prominent rhythmic features like tala, tempo from a polyphonic composition. For Western music, lots of work have been reported for automated drum analysis of polyphonic composition. However, attempts at computational analysis of tala by separating the tabla-signal from mixed signal in NIMS have not been successful. Tabla is played with two components - right and left. The right-hand component has frequency overlap with voice and other instruments. So, tala analysis of polyphonic-composition, by accurately separating the tabla-signal from the mixture is a baffling task, therefore an area of challenge. In this work we propose a novel technique for successfully detecting tala using left-tabla signal, producing meaningful results because the left-tabla normally doesn't have frequency overlap with voice and other instruments. North-Indian-rhythm follows complex cyclic pattern, against linear approach of Western-rhythm. We have exploited this cyclic property along with stressed and non-stressed methods of playing tabla-strokes to extract a characteristic pattern from the left-tabla strokes, which, after matching with the grammar of tala-system, determines the tala and tempo of the composition. A large number of polyphonic(vocal+tabla+other-instruments) compositions has been analyzed with the methodology and the result clearly reveals the effectiveness of proposed techniques.
△ Less
Submitted 24 September, 2018; v1 submitted 16 November, 2016;
originally announced November 2016.
