Surrogate-based Analysis Research Articles

Wind-Tunnel-in-the-Loop Exploration and Optimization of Active Flow Control Parameters

This paper considers the use of surrogate-based analysis and optimization (SBAO) methods to investigate the performance of pulsed jet actuators for active separation control in a wind tunnel. Two experimental setups are examined: pressure-induced separation on a one-sided diffuser and trailing-edge separation on a NACA 64A-015 airfoil. In both cases the modeling is done using Gaussian process regression (kriging), and the investigated active-flow-control parameters are the amplitude, frequency, and duty cycle of the actuators that are used to mitigate boundary-layer separation. In the diffuser test case, a parameter-space exploration is conducted to examine the effect of the three input parameters on the amount of reverse flow detected by an array of calorimetric shear-stress sensors. In the airfoil test case, an optimization strategy is followed to maximize an objective function constructed with the airfoil sectional lift coefficient and the mass flow consumption of the actuators. Both experiments consistently indicate that lowering the duty cycle of the pulsed-jet actuators below 0.5 may lead to efficiency gains in active separation control by limiting their mass flow consumption for equal performance, but with a concomitant supply pressure increase. Overall, the results presented herein demonstrate that SBAO methods could provide a potential for more efficient wind tunnel investigations involving multiparameter problems.

Surrogate Optimal Fractional Control for Constrained Operational Service of UAV Systems

In the expeditiously evolving discipline of autonomous aerial robotics, the efficiency and precision of drone control deliveries have become predominant. Different control strategies for UAV systems have been thoroughly investigated, yet PID controllers still receive significant consideration at various levels in the control loop. Although fractional-order PID controllers (FOPID) have greater flexibility than integer-order PID (IOPID) controllers, they are approached with caution and hesitance. This is due to the fact that FOPID controllers are more computationally intensive to tune, as well as being more challenging to implement accurately in real time. In this paper, we address this problem by developing and implementing a surrogate-based analysis and optimization (SBAO) of a relatively high-order approximation of FOPID controllers. The proposed approach was verified through two case studies; a simulation quadrotor benchmark model for waypoint navigation, and a real-time twin-rotor copter system. The obtained results validated and favored the SBAO approach over other classical heuristic methods for IOPID and FOPID.

Surrogate Model Based Analysis of Inter-Ply Shear Stress in Fiber Reinforced Thermoplastic Composite Sheet Press Forming

The anisotropic nature of fiber reinforced composite materials causes great challenges in predicting the inter-ply shear stress during forming. The complexity of understanding the functional dependency of inter-ply shear stress on multiple forming parameters such as blank temperature, pressure load, inter-ply slippage, and the relative fiber orientation angle of adjacent plies further limits the effort to produce a defect-free composite structure. Performing real experiments for various combinations of the mentioned parameters is both time consuming and economically costly. To overcome these difficulties, a surrogate-based analysis of inter-ply shear stress is proposed in this study. Based on the ranges of the forming parameters, computer experiments were performed. Using these experimental data, a radial basis function (RBF) based surrogate model that mimics inter-ply shear stress during composite press forming was developed. The fidelity of this model was checked with test data and found to be over 98% efficient.

Surrogate-Based Analysis and Design Optimization of Power Delivery Networks

As microprocessor architectures continue to increase computing performance under low-energy consumption, the combination of signal integrity, electromagnetic interference, and power delivery (PD) is becoming crucial in the computer industry. In that context, PD engineers make use of complex and computationally expensive models that impose time-consuming industrial practices to reach an adequate PD design. In this article, we propose a general surrogate-based methodology for fast and reliable analysis and design optimization of PD networks (PDN). We first formulate a generic surrogate model methodology exploiting passive lumped models optimized by parameter extraction to fit PDN impedance profiles. This PDN modeling formulation is illustrated with industrial laboratory measurements of a fourth generation server CPU motherboard. We next propose a black box PDN surrogate modeling methodology for efficient and reliable PD design optimization. To build our black box PDN surrogate, we compare four metamodeling techniques: support vector machines, polynomial surrogate modeling, generalized regression neural networks, and Kriging. The resultant best metamodel is then used to enable fast and accurate optimization of the PDN performance. Two examples validate our surrogate-based optimization approach: a voltage regulator (VR) with dual power rail remote sensing intended for communications and storage applications, by finding optimal sensing resistors and loading conditions; and a multiphase VR from a fifth-generation Intel server motherboard, by finding optimal compensation settings to reduce the number of bulk capacitors without losing CPU performance.

Handling model complexity with parsimony: Numerical analysis of the nitrogen turnover in a controlled aquifer model setup

Highly parameterized numerical models are frequently used in environmental sciences to interpret physicochemical processes. Despite their popularity, unjustified model complexity often undermines model generalizability, resulting in poor model predictive capabilities. A trade-off between model performance and complexity can be achieved using model selection techniques, which are largely unexplored in biogeochemical modeling. Our study aims to address this scientific gap by investigating the main advantages and findings of model selection in a typical biogeochemical modeling application. In particular, the study focuses on the numerical description, at five different levels of complexity, of the reactive transport processes of nitrogen species in a controlled aquifer model laboratory setup. To this end, the mechanistic model HYDRUS is coupled with the multimodal Nested Sampling algorithm for the estimation of the Bayesian model evidence and posterior parameter distributions. After successfully calibrating the HYDRUS-2D model against measurements from a tracer test, a lower-fidelity surrogate-based analysis is carried out to tackle the problem of computational cost, as well as to compare five different models which combine first-order and Monod-based parameterization of the microbial nitrogen degradation. Results demonstrate that measured data contain enough information content to support an increase in the complexity of the Monod model, which can reproduce the transient accumulation of nitrite at the beginning of the experiment. The Bayesian analysis reveals that such behavior is related to the particular structure of measured data, which plays a fundamental role in the model selection. The analysis concludes with a validation of the surrogate-based analysis, which confirms the reliability of the proposed numerical approach, opening up new perspectives on the use of multi-fidelity surrogates for the Bayesian model selection in environmental modeling.

Numerical simulations and surrogate-based optimization of cavitation performance for an aviation fuel pump

We used computational modeling to investigate the cavitation performance of an aviation fuel pump, and optimize structural parameters using the surrogate-based method. In the numerical simulation, a rotation-curvature correction was adapted to the k-e turbulence model, and a four-component surrogate fuel was selected to reproduce the physical properties of the China RP-3 kerosene. Then the performance of the aviation fuel pump was predicted. In the optimization, based on the series of the numerical results, Surrogate-based analysis and optimization (SBAO) was used to optimize the structural parameters of the fuel pump (the variation of the outlet blade angle for the inducer △βb1 and the variation of the inlet blade angle for the impeller △βb2). The results show that the prediction of cavitation performance agrees well with the experimental data. The results show that cavitation areas are mainly distributed in the inlet of the inducer. The volume of cavities grows with the decreasing NPSHa. The head of the fuel pump has a sudden head-drop when NPSHa ≤ 5.64 m. Furthermore, the surrogate-based approach is available in structural optimization of the fuel pump. The cavitation performance of the optimized pump improved about 22 % with a little drop of head coefficient when △βb1 = 4.33° and △βb2 = 3.24°. The numerical approach employed in this paper can accurately predict the cavitating flow of the high rotating speed fuel pump and the surrogate-based method is available in the structural optimization for a better cavitation performance.

Surrogate-Based Analysis and Optimization for the Design of Heat Sinks With Jet Impingement

Heat sinks are widely used for cooling electronic devices and systems. Their thermal performance is usually determined by the material, shape, and size of the heat sink. With the assistance of computational fluid dynamics (CFD) and surrogate-based optimization, heat sinks can be designed and optimized to achieve a high level of performance. In this paper, the design and optimization of a plate-fin-type heat sink cooled by impingement jet is presented. The flow and thermal fields are simulated using the CFD simulation; the thermal resistance of the heat sink is then estimated. A Kriging surrogate model is developed to approximate the objective function (thermal resistance) as a function of design variables. Surrogate-based optimization is implemented by adaptively adding infill points based on an integrated strategy of the minimum value, the maximum mean square error approach, and the expected improvement approaches. The results show the influence of design variables on the thermal resistance and give the optimal heat sink with lowest thermal resistance for given jet impingement conditions.

Optimization of LiMn2O4 electrode properties in a gradient- and surrogate-based framework

In this study, the effects of discharge rate and LiMn2O4 cathode properties (thickness, porosity, particle size, and solid-state diffusivity and conductivity) on the gravimetric energy and power density of a lithium-ion battery cell are analyzed simultaneously using a cell-level model. Surrogate-based analysis tools are applied to simulation data to construct educed-order models, which are in turn used to perform global sensitivity analysis to compare the relative importance of cathode properties. Based on these results, the cell is then optimized for several distinct physical scenarios using gradient-based methods. The complementary nature of the gradient- and surrogate-based tools is demonstrated by establishing proper bounds and constraints with the surrogate model, and then obtaining accurate optimized solutions with the gradient-based optimizer. These optimal solutions enable the quantification of the tradeoffs between energy and power density, and the effect of optimizing the electrode thickness and porosity. In conjunction with known guidelines, the numerical optimization framework developed herein can be applied directly to cell and pack design.

Alternative Cokriging Method for Variable-Fidelity Surrogate Modeling

Surrogate modeling plays an increasingly important role in different areas of aerospace engineering, such as erodynamic shape optimization, aerodynamic data production, structural design, and multidisciplinary design optimization of aircraft or spacecraft. Cokriging provides an attractive alternative approach to conventional kriging to improve the efficiency of building a surrogate model. It was initially proposed and applied in the geostatistics community for the enhanced prediction of less intensively sampled primary variables of interest with the assistance of intensively sampled auxiliary variables. As the underlying theory of cokriging is that of two-variable or multivariable kriging, it can be regarded as a general extension of (one-variable) kriging to a model that is assisted by auxiliary variables or secondary information. In an attempt to apply cokriging to the surrogate modeling problems associated with deterministic computer experiments, this article is motivated by the development of an alternative cokriging method to address the challenge related to the construction of the covariance matrix of cokriging [7]. Earlier work done by other authors related to this study can be found in the statistical community. For example, Kennedy and O’Hagan (KOH) proposed an autoregressive model to calculate the covariances and crosscovariances in the covariance matrix and developed a Bayesian approach to predict the output from an expensive high-fidelity simulation code with the assistance of lower-fidelity simulation codes. This Bayesian approach is identical to a form of cokriging suitable for computer experiments. Later, Qian andWu proposed a similar method, in which a random function (Gaussian process model) was used to replace the constant multiplicative factor of KOH’s method to account for the nonlinear scale change. KOH’s method was applied to multifidelity analysis and design optimization in the context of aerospace engineering by Forrester et al. and Kuya et al. More recently, Zimmerman and Han proposed a cokriging method with simplified cross-correlation estimation. In this article, we propose an alternative approach for the construction of the cokriging covariance matrix and develop a more practical cokriging method in the context of surrogate-based analysis and optimization. The developed cokriging method is validated against an analytical problem and applied to construct global approximation models of the aerodynamic coefficients as well as the drag polar of an RAE 2822 airfoil.

Surrogate-based modeling and dimension reduction techniques for multi-scale mechanics problems

Successful modeling and/or design of engineering systems often requires one to address the impact of multiple “design variables” on the prescribed outcome. There are often multiple, competing objectives based on which we assess the outcome of optimization. Since accurate, high fidelity models are typically time consuming and computationally expensive, comprehensive evaluations can be conducted only if an efficient framework is available. Furthermore, informed decisions of the model/hardware’s overall performance rely on an adequate understanding of the global, not local, sensitivity of the individual design variables on the objectives. The surrogate-based approach, which involves approximating the objectives as continuous functions of design variables from limited data, offers a rational framework to reduce the number of important input variables, i.e., the dimension of a design or modeling space. In this paper, we review the fundamental issues that arise in surrogate-based analysis and optimization, highlighting concepts, methods, techniques, as well as modeling implications for mechanics problems. To aid the discussions of the issues involved, we summarize recent efforts in investigating cryogenic cavitating flows, active flow control based on dielectric barrier discharge concepts, and lithium (Li)-ion batteries. It is also stressed that many multi-scale mechanics problems can naturally benefit from the surrogate approach for “scale bridging.”

Effective Transport Properties of LiMn2O4 Electrode via Particle-Scale Modeling

The extension of Li-ion batteries, from portable electronics to hybrid and electric vehicles, is significant. Developing a better understanding of the role of material properties and manipulating the morphology of the particle clusters comprising Li-ion electrodes could lead to potential opportunities for attaining higher performance goals, for which the effect of both material properties and morphology needs to be considered in a physics-based model. In this work, different particle packing arrangements are analyzed for the calculation of effective transport properties and reaction density that appear in the porous-electrode formulation due to the volume-averaging process. Surrogate-based analysis is used to systematically construct and validate reduced-order models for species transport at the particle-electrolyte interface. The low effective solid transport predicted through microscale modeling indicates the effect of packing arrangement and tortuosity, an aspect not captured by the Bruggeman’s relation. Particle cluster simulations reveal a Li-ion flux quantitatively different than that predicted by the porous-electrode model due to the variation of overpotential at the microscale. The present study offers a first-step towards integration of the effect of microstructure into a macroscale simulation.

Practically trivial parallel data processing gives neuroscience laboratories easy access to advanced analysis methods

Event Abstract Back to Event Practically trivial parallel data processing gives neuroscience laboratories easy access to advanced analysis methods Michael Denker1*, Bernd Wiebelt2, Denny Fliegner3, Markus Diesmann1 and Abigail Morrison2 1 RIKEN Brain Science Institute, Japan 2 Bernstein Center for Computational Neuroscience, Germany 3 Max Planck Institute for Dynamics and Self-Organization , Germany In addition to the increasing amounts of data gushing out from neuroscientific experiments, the complexity of modern data analysis techniques places new demands on the computing infrastructure required for data processing. In particular, the observation that neuronal data typically exhibit non-stationary statistics complicates the task of finding the correct null-hypothesis to assess the significance of a variety of test parameters. Modern computer resources enable a data-based approach to tackle significance estimation: surrogate techniques. In this framework the original data is modified in a specific way so as to keep some aspects of the data (e.g., the non-stationary nature of the data), while deliberately destroying others (i.e., those described by the test parameter). Repeating this procedure many times estimates the distribution of the test parameter under the null hypothesis. However, the required resources exceed the speed and memory constraints of a classical serial program design and require scientists to parallelize their analysis processes on distributed computer systems. Here, we explore step-by-step how to transform on-the-fly a typical data analysis program into a parallelized application. This approach is facilitated by the observation that a typical task in neuronal data analysis constitutes an embarrassingly parallel problem: the analysis can be divided up into independent parts that can be computed in parallel without communication. In particular for surrogate-based analysis programs, finding the decomposition of the analysis program into independent components is often trivial due to the inherent repetition of analysis steps. On the conceptual level, we demonstrate how in general to identify those parts of a serial program best suited for parallel execution. On the level of the practical implementation, we introduce four methods that assist in managing and distributing the parallelized code. By combining readily available high-level scientific programming languages and techniques for job control with metaprogramming no knowledge of system-level parallelization and the hardware architecture is required. We describe the solutions in a general fashion to facilitate the transfer of insights to the specific software and operating system environment of a particular laboratory. The details of our technique accompanied by concrete examples form a chapter of the new book “Analysis of parallel spike trains” edited by Sonja Grün and Stefan Rotter and published at Springer 2010. Conference: Neuroinformatics 2010 , Kobe, Japan, 30 Aug - 1 Sep, 2010. Presentation Type: Poster Presentation Topic: Electrophysiology Citation: Denker M, Wiebelt B, Fliegner D, Diesmann M and Morrison A (2010). Practically trivial parallel data processing gives neuroscience laboratories easy access to advanced analysis methods. Front. Neurosci. Conference Abstract: Neuroinformatics 2010 . doi: 10.3389/conf.fnins.2010.13.00110 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 15 Jun 2010; Published Online: 15 Jun 2010. * Correspondence: Michael Denker, RIKEN Brain Science Institute, Wako, Japan, mdenker@brain.riken.jp Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract The Authors in Frontiers Michael Denker Bernd Wiebelt Denny Fliegner Markus Diesmann Abigail Morrison Google Michael Denker Bernd Wiebelt Denny Fliegner Markus Diesmann Abigail Morrison Google Scholar Michael Denker Bernd Wiebelt Denny Fliegner Markus Diesmann Abigail Morrison PubMed Michael Denker Bernd Wiebelt Denny Fliegner Markus Diesmann Abigail Morrison Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.

Intercalation-Induced Stress and Heat Generation within Single Lithium-Ion Battery Cathode Particles

Intercalation-induced stress and heat generation inside Li-ion battery cathode particles under potentiodynamic control are simulated in this paper. We combined analyses of transport and kinetics in determining resulting stresses, which arise from concentration gradients in cathode particles, and heat generation. Two peaks in boundary reaction flux, and resulting stresses, were determined from the modeling of electrochemical kinetics and diffusion, using intrinsic material properties (resulting in two plateaus in the open-circuit potential) and the applied potential. Resistive heating was identified as the most important heat generation source. To probe the impact of the particle shape (equivalent radius and aspect ratio of an ellipsoidal particle) and the potential sweep rate on stress and heat generation, a surrogate-based analysis was also conducted. The systematic study showed that both intercalation-induced stress and time-averaged resistive heat generation rate increase with particle radius and potential sweep rate. Intercalation-induced stress increases first, then decreases as the aspect ratio of an ellipsoidal particle increases, whereas time-averaged resistive heat generation rate decreases as aspect ratio increases. This surrogate-based analysis suggests that ellipsoidal particles with larger aspect ratios are preferred over spherical particles, in improving battery performance when stress and heat generation are the only factors considered.