scholarly journals Bayesian optimization with adaptive surrogate models for automated experimental design

2021 ◽  
Vol 7 (1) ◽  
Author(s):  
Bowen Lei ◽  
Tanner Quinn Kirk ◽  
Anirban Bhattacharya ◽  
Debdeep Pati ◽  
Xiaoning Qian ◽  
...  
Keyword(s):  
Experimental Design ◽  
Design Space ◽  
Materials Science ◽  
Scientific Discovery ◽  
Surrogate Models ◽  
Multivariate Adaptive Regression Splines ◽  
Bayesian Optimization ◽  
Objective Functions ◽  
Additive Regression ◽  
Efficient Exploration

AbstractBayesian optimization (BO) is an indispensable tool to optimize objective functions that either do not have known functional forms or are expensive to evaluate. Currently, optimal experimental design is always conducted within the workflow of BO leading to more efficient exploration of the design space compared to traditional strategies. This can have a significant impact on modern scientific discovery, in particular autonomous materials discovery, which can be viewed as an optimization problem aimed at looking for the maximum (or minimum) point for the desired materials properties. The performance of BO-based experimental design depends not only on the adopted acquisition function but also on the surrogate models that help to approximate underlying objective functions. In this paper, we propose a fully autonomous experimental design framework that uses more adaptive and flexible Bayesian surrogate models in a BO procedure, namely Bayesian multivariate adaptive regression splines and Bayesian additive regression trees. They can overcome the weaknesses of widely used Gaussian process-based methods when faced with relatively high-dimensional design space or non-smooth patterns of objective functions. Both simulation studies and real-world materials science case studies demonstrate their enhanced search efficiency and robustness.

scholarly journals Design of Mechanical Metamaterials via Constrained Bayesian Optimization

2018 ◽  
Author(s):  
Conner Sharpe ◽  
Carolyn Conner Seepersad ◽  
Seth Watts ◽  
Dan Tortorelli
Keyword(s):  
Design Space ◽  
Engineering Problem ◽  
Bayesian Optimization ◽  
Probabilistic Constraints ◽  
Objective Functions ◽  
Bulk Properties ◽  
New Class ◽  
Naturally Occurring ◽  
Design Variables ◽  
Mechanical Metamaterials

Advances in additive manufacturing processes have made it possible to build mechanical metamaterials with bulk properties that exceed those of naturally occurring materials. One class of these metamaterials is structural lattices that can achieve high stiffness to weight ratios. Recent work on geometric projection approaches has introduced the possibility of optimizing these architected lattice designs in a drastically reduced parameter space. The reduced number of design variables enables application of a new class of methods for exploring the design space. This work investigates the use of Bayesian optimization, a technique for global optimization of expensive non-convex objective functions through surrogate modeling. We utilize formulations for implementing probabilistic constraints in Bayesian optimization to aid convergence in this highly constrained engineering problem, and demonstrate results with a variety of stiff lightweight lattice designs.

Enhanced Multi-Objective Optimization of a Microchannel Heat Sink Using Multiple Surrogates Modeling

2009 ◽  
Author(s):  
Afzal Husain ◽  
Kwang-Yong Kim
Keyword(s):  
Heat Sink ◽  
Design Space ◽  
Surrogate Models ◽  
Pareto Optimal ◽  
Microchannel Heat Sink ◽  
Objective Functions ◽  
Multi Objective Optimization ◽  
Pareto Optimal Front ◽  
Multi Objective ◽  
Design Variables

A liquid flow microchannel heat sink has been studied and optimized with the help of three-dimensional numerical analysis and multiple surrogate methods. Two objective functions, thermal resistance and pumping power have been selected to assess the performance of the microchannel heat sink. The design variables related to the microchannel top and bottom widths, depth and fin width, which contribute to objective functions, have been identified and design space has been explored through some preliminary calculations. Design of experiments was performed and a three-level full factorial design was selected to exploit the design space. The numerical solutions obtained at these design points were utilized to construct surrogate models namely Response Surface Approximations and Kriging. A hybrid multi-objective evolutionary algorithm coupled with surrogate models and a gradient-based search algorithm is applied to find global Pareto-optimal solutions. Since, the surrogate models are highly problem-dependent, the accuracy of the two surrogate models has been discussed in view of their predictions at on- and off-Pareto-optimal front. The trade-off analysis was performed in view of the two competing objectives. The Pareto-optimal sensitivity (change in value along the Pareto-optimal front) of the design variables has been found out to economically compromise with the design variables contributing relatively less to the objective functions. The application of the multiple surrogate methods not only improves quality of multi-objective optimization but also gives the feedback of the fidelity of the model near the optimum region.

scholarly journals Two-step machine learning enables optimized nanoparticle synthesis

2021 ◽  
Vol 7 (1) ◽  
Author(s):  
Flore Mekki-Berrada ◽  
Zekun Ren ◽  
Tan Huang ◽  
Wai Kuan Wong ◽  
Fang Zheng ◽  
...  
Keyword(s):  
Machine Learning ◽  
Optical Properties ◽  
Materials Science ◽  
Nanoparticle Synthesis ◽  
Bayesian Optimization ◽  
Absorbance Spectrum ◽  
Nanomaterial Synthesis ◽  
Target Spectrum ◽  
Colour Palette ◽  
Algorithmic Framework

AbstractIn materials science, the discovery of recipes that yield nanomaterials with defined optical properties is costly and time-consuming. In this study, we present a two-step framework for a machine learning-driven high-throughput microfluidic platform to rapidly produce silver nanoparticles with the desired absorbance spectrum. Combining a Gaussian process-based Bayesian optimization (BO) with a deep neural network (DNN), the algorithmic framework is able to converge towards the target spectrum after sampling 120 conditions. Once the dataset is large enough to train the DNN with sufficient accuracy in the region of the target spectrum, the DNN is used to predict the colour palette accessible with the reaction synthesis. While remaining interpretable by humans, the proposed framework efficiently optimizes the nanomaterial synthesis and can extract fundamental knowledge of the relationship between chemical composition and optical properties, such as the role of each reactant on the shape and amplitude of the absorbance spectrum.

Optimal experimental design with fast neural network surrogate models

2021 ◽  
Vol 200 ◽  
pp. 110747
Author(s):  
Joshua Stuckner ◽  
Matthew Piekenbrock ◽  
Steven M. Arnold ◽  
Trenton M. Ricks
Keyword(s):  
Neural Network ◽  
Experimental Design ◽  
Surrogate Models ◽  
Optimal Experimental Design

Stochastic Bayesian optimization for predicting borderline knock

2021 ◽  
pp. 146808742110652
Author(s):  
Jian Tang ◽  
Anuj Pal ◽  
Wen Dai ◽  
Chad Archer ◽  
James Yi ◽  
...  
Keyword(s):  
Fuel Economy ◽  
Learning Algorithm ◽  
Feedforward Control ◽  
Model Development ◽  
Kriging Model ◽  
Optimization Method ◽  
Surrogate Models ◽  
Bayesian Optimization ◽  
Control Parameters ◽  
Operation Conditions

Engine knock is an undesirable combustion that could damage the engine mechanically. On the other hand, it is often desired to operate the engine close to its borderline knock limit to optimize combustion efficiency. Traditionally, borderline knock limit is detected by sweeping tests of related control parameters for the worst knock, which is expensive and time consuming, and also, the detected borderline knock limit is often used as a feedforward control without considering its stochastic characteristics without compensating current engine operational condition and type of fuel used. In this paper, stochastic Bayesian optimization method is used to obtain a tradeoff between stochastic knock intensity and fuel economy. The log-nominal distribution of knock intensity signal is converted to Gaussian one using a proposed map to satisfy the assumption for Kriging model development. Both deterministic and stochastic Kriging surrogate models are developed based on test data using the Bayesian iterative optimization process. This study focuses on optimizing two competing objectives, knock intensity and indicated specific fuel consumption using two control parameters: spark and intake valve timings. Test results at two different operation conditions show that the proposed learning algorithm not only reduces required time and cost for predicting knock borderline but also provides control parameters, based on trained surrogate models and the corresponding Pareto front, with the best fuel economy possible.

On Adaptive Sampling for Single and Multi-Response Bayesian Surrogate Models

2006 ◽  
Author(s):  
David A. Romero ◽  
Cristina H. Amon ◽  
Susan Finger
Keyword(s):  
Experimental Design ◽  
Adaptive Sampling ◽  
Design Space Exploration ◽  
A Priori ◽  
Computer Experiments ◽  
Latin Hypercube Sampling ◽  
Surrogate Models ◽  
Sampling Techniques ◽  
Design And Optimization ◽  
Sampling Points

In order to reduce the time and resources devoted to design-space exploration during simulation-based design and optimization, the use of surrogate models, or metamodels, has been proposed in the literature. Key to the success of metamodeling efforts are the experimental design techniques used to generate the combinations of input variables at which the computer experiments are conducted. Several adaptive sampling techniques have been proposed to tailor the experimental designs to the specific application at hand, using the already-acquired data to guide further exploration of the input space, instead of using a fixed sampling scheme defined a priori. Though mixed results have been reported, it has been argued that adaptive sampling techniques can be more efficient, yielding better surrogate models with less sampling points. In this paper, we address the problem of adaptive sampling for single and multi-response metamodels, with a focus on Multi-stage Multi-response Bayesian Surrogate Models (MMBSM). We compare distance-optimal latin hypercube sampling, an entropy-based criterion and the maximum cross-validation variance criterion, originally proposed for one-dimensional output spaces and implemented in this paper for multi-dimensional output spaces. Our results indicate that, both for single and multi-response surrogate models, the entropy-based adaptive sampling approach leads to models that are more robust to the initial experimental design and at least as accurate (or better) when compared with other sampling techniques using the same number of sampling points.

Towards the Multiobjective Optimisation of Gas Turbine Combustors

2006 ◽  
Author(s):  
S. G. Wyse ◽  
G. T. Parks ◽  
R. S. Cant
Keyword(s):  
Transfer Function ◽  
Tabu Search ◽  
Combustion Chamber ◽  
Gas Turbine ◽  
Design Space ◽  
Objective Functions ◽  
Multiobjective Optimisation ◽  
Gas Turbine Combustor ◽  
Linear Network ◽  
Good Design

Gas turbine combustor design entails multiple, and often contradictory, requirements for the designer to consider. Multiobjective optimisation on a low-fidelity linear-network-based code is suggested as a way of investigating the design space. The ability of the Tabu Search optimiser to minimise NOx and CO, as well as several acoustic objective functions, is investigated, and the resulting “good” design vectors presented. An analysis of the importance of the flame transfer function in the model is also given. The mass flow and the combustion chamber width and area are shown to be very important. The length of the plenum and the widths of the plenum exit and combustor exit also influence the design space.

scholarly journals Experimental Design for Optimization of Orthogonal Projection Pursuit Models

2020 ◽  
Vol 34 (06) ◽  
pp. 10235-10242
Author(s):  
Mojmir Mutny ◽  
Johannes Kirschner ◽  
Andreas Krause
Keyword(s):  
Experimental Design ◽  
Orthogonal Projection ◽  
Optimization Problems ◽  
Additive Model ◽  
Projection Pursuit ◽  
Parameter Tuning ◽  
Additive Models ◽  
Function Optimization ◽  
Bayesian Optimization ◽  
Additive Structure

Bayesian optimization and kernelized bandit algorithms are widely used techniques for sequential black box function optimization with applications in parameter tuning, control, robotics among many others. To be effective in high dimensional settings, previous approaches make additional assumptions, for example on low-dimensional subspaces or an additive structure. In this work, we go beyond the additivity assumption and use an orthogonal projection pursuit regression model, which strictly generalizes additive models. We present a two-stage algorithm motivated by experimental design to first decorrelate the additive components. Subsequently, the bandit optimization benefits from the statistically efficient additive model. Our method provably decorrelates the fully additive model and achieves optimal sublinear simple regret in terms of the number of function evaluations. To prove the rotation recovery, we derive novel concentration inequalities for linear regression on subspaces. In addition, we specifically address the issue of acquisition function optimization and present two domain dependent efficient algorithms. We validate the algorithm numerically on synthetic as well as real-world optimization problems.

Sign in / Sign up

Export Citation Format

Share Document