Machine Learning to Aid Tuning of Numerical Parameters in Topology Optimization

2019 ◽  
Vol 141 (11) ◽  
Author(s):  
Matthew E. Lynch ◽  
Soumalya Sarkar ◽  
Kurt Maute
Keyword(s):  
Machine Learning ◽  
Topology Optimization ◽  
Optimization Problems ◽  
Bayesian Optimization ◽  
Computational Time ◽  
Optimization Approach ◽  
Learning Approaches ◽  
Solution Quality ◽  
Topology Optimization Problem ◽  
Tuning Parameters

Abstract Recent advances in design optimization have significant potential to improve the function of mechanical components and systems. Coupled with additive manufacturing, topology optimization is one category of numerical methods used to produce algorithmically generated optimized designs making a difference in the mechanical design of hardware currently being introduced to the market. Unfortunately, many of these algorithms require extensive manual setup and control, particularly of tuning parameters that control algorithmic function and convergence. This paper introduces a framework based on machine learning approaches to recommend tuning parameters to a user in order to avoid costly trial and error involved in manual tuning. The algorithm reads tuning parameters from a repository of prior, similar problems adjudged using a dissimilarity metric based on problem metadata and refines them for the current problem using a Bayesian optimization approach. The approach is demonstrated for a simple topology optimization problem with the objective of achieving good topology optimization solution quality and then with the additional objective of finding an optimal “trade” between solution quality and required computational time. The goal is to reduce the total number of “wasted” tuning runs that would be required for purely manual tuning. With more development, the framework may ultimately be useful on an enterprise level for analysis and optimization problems—topology optimization is one example but the framework is also applicable to other optimization problems such as shape and sizing and in high-fidelity physics-based analysis models—and enable these types of advanced approaches to be used more efficiently.

Multifidelity and Multiscale Bayesian Framework for High-Dimensional Engineering Design and Calibration

2019 ◽  
Vol 141 (12) ◽  
Author(s):  
Soumalya Sarkar ◽  
Sudeepta Mondal ◽  
Michael Joly ◽  
Matthew E. Lynch ◽  
Shaunak D. Bopardikar ◽  
...  
Keyword(s):  
Machine Learning ◽  
Learning Strategies ◽  
Optimization Problems ◽  
Bayesian Optimization ◽  
High Dimensional ◽  
Optimization Approach ◽  
Information Theoretic ◽  
Compressor Rotor ◽  
Practical Engineering ◽  
Microstructure Prediction

AbstractThis paper proposes a machine learning–based multifidelity modeling (MFM) and information-theoretic Bayesian optimization approach where the associated models can have complex discrepancies among each other. Advantages of MFM-based optimization over a single-fidelity surrogate, specifically under complex constraints, are discussed with benchmark optimization problems involving noisy data. The MFM framework, based on modeling of the varied fidelity information sources via Gaussian processes, is augmented with information-theoretic active learning strategies that involve sequential selection of optimal points in a multiscale architecture. This framework is demonstrated to exhibit improved efficiency on practical engineering problems like high-dimensional design optimization of compressor rotor via implementing its multiscale architecture and calibration of expensive microstructure prediction model. From the perspective of the machine learning–assisted design of multiphysics systems, advantages of the proposed framework have been presented with respect to accelerating the search of optimal design conditions under budget constraints.

scholarly journals Practical CO2—WAG Field Operational Designs Using Hybrid Numerical-Machine-Learning Approaches

Energies ◽  
2021 ◽  
Vol 14 (4) ◽  
pp. 1055
Author(s):  
Qian Sun ◽  
William Ampomah ◽  
Junyu You ◽  
Martha Cather ◽  
Robert Balch
Keyword(s):  
Machine Learning ◽  
Oil Recovery ◽  
History Matching ◽  
Optimization Problems ◽  
Learning Technologies ◽  
Petroleum Engineering ◽  
Support Vector ◽  
Learning Approaches ◽  
Field Development ◽  
Proxy Models

Machine-learning technologies have exhibited robust competences in solving many petroleum engineering problems. The accurate predictivity and fast computational speed enable a large volume of time-consuming engineering processes such as history-matching and field development optimization. The Southwest Regional Partnership on Carbon Sequestration (SWP) project desires rigorous history-matching and multi-objective optimization processes, which fits the superiorities of the machine-learning approaches. Although the machine-learning proxy models are trained and validated before imposing to solve practical problems, the error margin would essentially introduce uncertainties to the results. In this paper, a hybrid numerical machine-learning workflow solving various optimization problems is presented. By coupling the expert machine-learning proxies with a global optimizer, the workflow successfully solves the history-matching and CO2 water alternative gas (WAG) design problem with low computational overheads. The history-matching work considers the heterogeneities of multiphase relative characteristics, and the CO2-WAG injection design takes multiple techno-economic objective functions into accounts. This work trained an expert response surface, a support vector machine, and a multi-layer neural network as proxy models to effectively learn the high-dimensional nonlinear data structure. The proposed workflow suggests revisiting the high-fidelity numerical simulator for validation purposes. The experience gained from this work would provide valuable guiding insights to similar CO2 enhanced oil recovery (EOR) projects.

scholarly journals Machine learning approaches to calibrate individual-based infectious disease models

2021 ◽  
Author(s):  
Theresa Reiker ◽  
Monica Golumbeanu ◽  
Andrew Shattock ◽  
Lydia Burgert ◽  
Thomas A. Smith ◽  
...  
Keyword(s):  
Machine Learning ◽  
Disease Transmission ◽  
Goodness Of Fit ◽  
Epidemiological Data ◽  
Bayesian Optimization ◽  
Model Complexity ◽  
Learning Approaches ◽  
Dimensional Parameter ◽  
Novel Approach ◽  
Transmission Models

AbstractIndividual-based models have become important tools in the global battle against infectious diseases, yet model complexity can make calibration to biological and epidemiological data challenging. We propose a novel approach to calibrate disease transmission models via a Bayesian optimization framework employing machine learning emulator functions to guide a global search over a multi-objective landscape. We demonstrate our approach by application to an established individual-based model of malaria, optimizing over a high-dimensional parameter space with respect to a portfolio of multiple fitting objectives built from datasets capturing the natural history of malaria transmission and disease progression. Outperforming other calibration methodologies, the new approach quickly reaches an improved final goodness of fit. Per-objective parameter importance and sensitivity diagnostics provided by our approach offer epidemiological insights and enhance trust in predictions through greater interpretability.One Sentence SummaryWe propose a novel, fast, machine learning-based approach to calibrate disease transmission models that outperforms other methodologies

Robust topology optimization for continuum structures with random loads

2018 ◽  
Vol 35 (2) ◽  
pp. 710-732 ◽  
Author(s):  
Jie Liu ◽  
Guilin Wen ◽  
Qixiang Qing ◽  
Fangyi Li ◽  
Yi Min Xie
Keyword(s):  
Topology Optimization ◽  
Optimization Method ◽  
Computational Time ◽  
Optimization Approach ◽  
Random Load ◽  
Content Type ◽  
Continuum Structures ◽  
Robust Topology Optimization ◽  
Random Loads ◽  
Expansion Technique

Purpose This paper aims to tackle the challenge topic of continuum structural layout in the presence of random loads and to develop an efficient robust method. Design/methodology/approach An innovative robust topology optimization approach for continuum structures with random applied loads is reported. Simultaneous minimization of the expectation and the variance of the structural compliance is performed. Uncertain load vectors are dealt with by using additional uncertain pseudo random load vectors. The sensitivity information of the robust objective function is obtained approximately by using the Taylor expansion technique. The design problem is solved using bi-directional evolutionary structural optimization method with the derived sensitivity numbers. Findings The numerical examples show the significant topological changes of the robust solutions compared with the equivalent deterministic solutions. Originality/value A simple yet efficient robust topology optimization approach for continuum structures with random applied loads is developed. The computational time scales linearly with the number of applied loads with uncertainty, which is very efficient when compared with Monte Carlo-based optimization method.

scholarly journals A Hadoop Based Framework Integrating Machine Learning Classifiers for Anomaly Detection in the Internet of Things

Electronics ◽  
2021 ◽  
Vol 10 (16) ◽  
pp. 1955
Author(s):  
Ikram Sumaiya Thaseen ◽  
Vanitha Mohanraj ◽  
Sakthivel Ramachandran ◽  
Kishore Sanapala ◽  
Sang-Soo Yeo
Keyword(s):  
Machine Learning ◽  
Internet Of Things ◽  
Experimental Analysis ◽  
Parameter Tuning ◽  
Computational Time ◽  
Support Vector ◽  
Learning Approaches ◽  
K Nearest Neighbor ◽  
Machine Learning Classifiers ◽  
Learning Classifiers

In recent years, different variants of the botnet are targeting government, private organizations and there is a crucial need to develop a robust framework for securing the IoT (Internet of Things) network. In this paper, a Hadoop based framework is proposed to identify the malicious IoT traffic using a modified Tomek-link under-sampling integrated with automated Hyper-parameter tuning of machine learning classifiers. The novelty of this paper is to utilize a big data platform for benchmark IoT datasets to minimize computational time. The IoT benchmark datasets are loaded in the Hadoop Distributed File System (HDFS) environment. Three machine learning approaches namely naive Bayes (NB), K-nearest neighbor (KNN), and support vector machine (SVM) are used for categorizing IoT traffic. Artificial immune network optimization is deployed during cross-validation to obtain the best classifier parameters. Experimental analysis is performed on the Hadoop platform. The average accuracy of 99% and 90% is obtained for BoT_IoT and ToN_IoT datasets. The accuracy difference in ToN-IoT dataset is due to the huge number of data samples captured at the edge layer and fog layer. However, in BoT-IoT dataset only 5% of the training and test samples from the complete dataset are considered for experimental analysis as released by the dataset developers. The overall accuracy is improved by 19% in comparison with state-of-the-art techniques. The computational times for the huge datasets are reduced by 3–4 hours through Map Reduce in HDFS.

scholarly journals MLGen: Generative Design Framework Based on Machine Learning and Topology Optimization

2021 ◽  
Vol 11 (24) ◽  
pp. 12044
Author(s):  
Nikos Ath. Kallioras ◽  
Nikos D. Lagaros
Keyword(s):  
Machine Learning ◽  
Topology Optimization ◽  
Architectural Design ◽  
Parametric Design ◽  
Optimization Problems ◽  
Design Procedure ◽  
Performance Criteria ◽  
Design Framework ◽  
Generative Design ◽  
New Era

Design and manufacturing processes are entering into a new era as novel methods and techniques are constantly introduced. Currently, 3D printing is already established in the production processes of several industries while more are continuously being added. At the same time, topology optimization has become part of the design procedure of various industries, such as automotive and aeronautical. Parametric design has been gaining ground in the architectural design literature in the past years. Generative design is introduced as the contemporary design process that relies on the utilization of algorithms for creating several forms that respect structural and architectural constraints imposed, among others, by the design codes and/or as defined by the designer. In this study, a novel generative design framework labeled as MLGen is presented. MLGen integrates machine learning into the generative design practice. MLGen is able to generate multiple optimized solutions which vary in shape but are equivalent in terms of performance criteria. The output of the proposed framework is exported in a format that can be handled by 3D printers. The ability of MLGen to efficiently handle different problems is validated via testing on several benchmark topology optimization problems frequently employed in the literature.

scholarly journals Development of a Recognition System for Spraying Areas from Unmanned Aerial Vehicles Using a Machine Learning Approach

Sensors ◽  
2019 ◽  
Vol 19 (2) ◽  
pp. 313 ◽  
Author(s):  
Pengbo Gao ◽  
Yan Zhang ◽  
Linhuan Zhang ◽  
Ryozo Noguchi ◽  
Tofael Ahamed
Keyword(s):  
Machine Learning ◽  
Field Experiments ◽  
Recognition Accuracy ◽  
Recognition System ◽  
Learning System ◽  
Computational Time ◽  
Learning Approaches ◽  
Online Systems ◽  
Online Recognition ◽  
Spray System

Unmanned aerial vehicle (UAV)-based spraying systems have recently become important for the precision application of pesticides, using machine learning approaches. Therefore, the objective of this research was to develop a machine learning system that has the advantages of high computational speed and good accuracy for recognizing spray and non-spray areas for UAV-based sprayers. A machine learning system was developed by using the mutual subspace method (MSM) for images collected from a UAV. Two target lands: agricultural croplands and orchard areas, were considered in building two classifiers for distinguishing spray and non-spray areas. The field experiments were conducted in target areas to train and test the system by using a commercial UAV (DJI Phantom 3 Pro) with an onboard 4K camera. The images were collected from low (5 m) and high (15 m) altitudes for croplands and orchards, respectively. The recognition system was divided into offline and online systems. In the offline recognition system, 74.4% accuracy was obtained for the classifiers in recognizing spray and non-spray areas for croplands. In the case of orchards, the average classifier recognition accuracy of spray and non-spray areas was 77%. On the other hand, the online recognition system performance had an average accuracy of 65.1% for croplands, and 75.1% for orchards. The computational time for the online recognition system was minimal, with an average of 0.0031 s for classifier recognition. The developed machine learning system had an average recognition accuracy of 70%, which can be implemented in an autonomous UAV spray system for recognizing spray and non-spray areas for real-time applications.

scholarly journals Biomarker Selection and Classification of “-Omics” Data Using a Two-Step Bayes Classification Framework

2013 ◽  
Vol 2013 ◽  
pp. 1-9 ◽  
Author(s):  
Anunchai Assawamakin ◽  
Supakit Prueksaaroon ◽  
Supasak Kulawonganunchai ◽  
Philip James Shaw ◽  
Vara Varavithya ◽  
...  
Keyword(s):  
Machine Learning ◽  
Naive Bayes ◽  
Naïve Bayes ◽  
Computational Time ◽  
Learning Approaches ◽  
Naive Bayes Classifier ◽  
Bayes Classifier ◽  
Naïve Bayes Classifier ◽  
Classification Framework ◽  
Hidden Naive Bayes

Identification of suitable biomarkers for accurate prediction of phenotypic outcomes is a goal for personalized medicine. However, current machine learning approaches are either too complex or perform poorly. Here, a novel two-step machine-learning framework is presented to address this need. First, a Naïve Bayes estimator is used to rank features from which the top-ranked will most likely contain the most informative features for prediction of the underlying biological classes. The top-ranked features are then used in a Hidden Naïve Bayes classifier to construct a classification prediction model from these filtered attributes. In order to obtain the minimum set of the most informative biomarkers, the bottom-ranked features are successively removed from the Naïve Bayes-filtered feature list one at a time, and the classification accuracy of the Hidden Naïve Bayes classifier is checked for each pruned feature set. The performance of the proposed two-step Bayes classification framework was tested on different types of -omicsdatasets including gene expression microarray, single nucleotide polymorphism microarray (SNParray), and surface-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) proteomic data. The proposed two-step Bayes classification framework was equal to and, in some cases, outperformed other classification methods in terms of prediction accuracy, minimum number of classification markers, and computational time.

A Non-Dominant Sorting Genetic Algorithm for Optimization of a Product Design and Selection of Its Suppliers

2020 ◽  
Vol 19 (01) ◽  
pp. 167-188
Author(s):  
Oulfa Labbi ◽  
Abdeslam Ahmadi ◽  
Latifa Ouzizi ◽  
Mohammed Douimi
Keyword(s):  
Genetic Algorithm ◽  
Product Design ◽  
Computational Time ◽  
Optimization Approach ◽  
Nsga Ii ◽  
Solution Quality ◽  
Holding Costs ◽  
Weighted Sum Method ◽  
The Impact ◽  
Selection Of

The aim of this paper is to address the problem of supplier selection in a context of an integrated product design. Indeed, the product specificities and the suppliers’ constraints are both integrated into product design phase. We consider the case of improving the design of an existing product and study the selection of its suppliers adopting a bi-objective optimization approach. Considering multi-products, multi-suppliers and multi-periods, the mathematical model proposed aims to minimize supplying, transport and holding costs of product components as well as quality rejected items. To solve the bi-objective problem, an evolutionary algorithm namely, non-dominant sorting genetic algorithm (NSGA-II) is employed. The algorithm provides a set of Pareto front solutions optimizing the two objective functions at once. Since parameters values of genetic algorithms have a significant impact on their efficiency, we have proposed to study the impact of each parameter on the fitness functions in order to determine the optimal combination of these parameters. Thus, a number of simulations evaluating the effects of crossover rate, mutation rate and number of generations on Pareto fronts are presented. To evaluate performance of the algorithm, results are compared to those obtained by the weighted sum method through a numerical experiment. According to the computational results, the non-dominant sorting genetic algorithm outperforms the CPLEX MIP solver in both solution quality and computational time.

scholarly journals Global Optimization for Mixed–Discrete Structural Design

Symmetry ◽  
2020 ◽  
Vol 12 (9) ◽  
pp. 1529
Author(s):  
Jung-Fa Tsai ◽  
Ming-Hua Lin ◽  
Duan-Yi Wen
Keyword(s):  
Structural Design ◽  
Design Problem ◽  
Optimization Problems ◽  
Global Optimum ◽  
Computational Time ◽  
Optimization Approach ◽  
Global Optimality ◽  
Discrete Variables ◽  
Beam Design ◽  
Continuous And Discrete Variables

Several structural design problems that involve continuous and discrete variables are very challenging because of the combinatorial and non-convex characteristics of the problems. Although the deterministic optimization approach theoretically guarantees to find the global optimum, it usually leads to a significant burden in computational time. This article studies the deterministic approach for globally solving mixed–discrete structural optimization problems. An improved method that symmetrically reduces the number of constraints for linearly expressing signomial terms with pure discrete variables is applied to significantly enhance the computational efficiency of obtaining the exact global optimum of the mixed–discrete structural design problem. Numerical experiments of solving the stepped cantilever beam design problem and the pressure vessel design problem are conducted to show the efficiency and effectiveness of the presented approach. Compared with existing methods, this study introduces fewer convex terms and constraints for transforming the mixed–discrete structural problem and uses much less computational time for solving the reformulated problem to global optimality.

Sign in / Sign up

Export Citation Format

Share Document