Enhanced Gaussian Process Metamodeling and Collaborative Optimization for Vehicle Suspension Design Optimization

Dynamic stability is a key performance metric of motor vehicles and has a direct impact on passenger experience and customer satisfaction. The desired vehicle dynamics behavior can be achieved by optimizing the design of vehicle suspensions. Two challenges are associated with this design optimization task. The first one arises from the large number (e.g., 40 or 50) of design variables in modern suspension systems. Such multitude of variables not only makes it expensive to build a training dataset for metamodeling purposes, but also renders accurate surrogate modeling extremely difficult. The second challenge is a lack of guideline for choosing a proper multidisciplinary design optimization (MDO) method for a single MDO problem such as one for vehicle suspension design. In this paper, an enhanced Gaussian process (GP) metamodeling technique is developed and several versions of the collaborative optimization (CO) method are compared via a vehicle suspension design problem. In our enhanced GP modeling method, the model parameters are efficiently estimated using the smoothing effect of the so-called nugget parameter to reduce the search space. In addition, various versions of the CO method are studied where the enhanced collaborative optimization (ECO) method is found to perform the best. A simplified ECO formulation is also investigated to provide insights for future engineering applications.

Download Full-text

Support Vector Regression-Based Multidisciplinary Design Optimization for Ship Design

Volume 1: Offshore Technology ◽

10.1115/omae2012-83086 ◽

2012 ◽

Author(s):

Dongqin Li ◽

Yifeng Guan ◽

Qingfeng Wang ◽

Zhitong Chen

Keyword(s):

Design Optimization ◽

Support Vector Regression ◽

Design Process ◽

Multidisciplinary Design Optimization ◽

Initial Point ◽

Ship Design ◽

Collaborative Optimization ◽

Multidisciplinary Design ◽

Support Vector ◽

Independent Design

The design of ship is related to several disciplines such as hydrostatic, resistance, propulsion and economic. The traditional design process of ship only involves independent design optimization within each discipline. With such an approach, there is no guarantee to achieve the optimum design. And at the same time improving the efficiency of ship optimization is also crucial for modem ship design. In this paper, an introduction of both the traditional ship design process and the fundamentals of Multidisciplinary Design Optimization (MDO) theory are presented and a comparison between the two methods is carried out. As one of the most frequently applied MDO methods, Collaborative Optimization (CO) promotes autonomy of disciplines while providing a coordinating mechanism guaranteeing progress toward an optimum and maintaining interdisciplinary compatibility. However there are some difficulties in applying the conventional CO method, such as difficulties in choosing an initial point and tremendous computational requirements. For the purpose of overcoming these problems, Design Of Experiment (DOE) and a new support vector regression algorithm are applied to CO to construct statistical approximation model in this paper. The support vector regression algorithm approximates the optimization model and is updated during the optimization process to improve accuracy. It is shown by examples that the computing efficiency and robustness of this CO method are higher than with the conventional CO method. Then this new Collaborative Optimization (CO) method using approximate technology is discussed in detail and applied in ship design which considers hydrostatic, propulsion, weight and volume, performance and cost. It indicates that CO method combined with approximate technology can effectively solve complex engineering design optimization problem. Finally, some suggestions on the future improvements are proposed.

Download Full-text

Study on Mix-Variable Collaborative Design Optimization

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.215-216.592 ◽

2012 ◽

Vol 215-216 ◽

pp. 592-596

Author(s):

Li Gao ◽

Rong Rong Wang

Keyword(s):

Design Optimization ◽

Product Design ◽

Optimization Algorithm ◽

Collaborative Design ◽

Optimization Problems ◽

Optimal Solution ◽

Collaborative Optimization ◽

Continuous Variables ◽

Complex Product ◽

Divide And Rule

In order to deal with complex product design optimization problems with both discrete and continuous variables, mix-variable collaborative design optimization algorithm is put forward based on collaborative optimization, which is an efficient way to solve mix-variable design optimization problems. On the rule of “divide and rule”, the algorithm decouples the problem into some relatively simple subsystems. Then by using collaborative mechanism, the optimal solution is obtained. Finally, the result of a case shows the feasibility and effectiveness of the new algorithm.

Download Full-text

Gaussian Process emulation of multi-model ice sheet and glacier projections

10.5194/egusphere-egu21-14561 ◽

2021 ◽

Author(s):

Tamsin Edwards ◽

Keyword(s):

Gaussian Process ◽

Sea Level Rise ◽

Sea Level ◽

Climate Models ◽

Ice Sheet ◽

Surface Air Temperature ◽

Model Parameters ◽

Gaussian Process Emulation ◽

Model Ensembles ◽

Global Mean

The land ice contribution to global mean sea level rise has not yet been predicted with ice sheet and glacier models for the latest set of socio-economic scenarios (SSPs), nor with coordinated exploration of uncertainties arising from the various computer models involved. Two recent international projects (ISMIP6 and GlacierMIP) generated a large suite of projections using multiple models, but mostly used previous generation scenarios and climate models, and could not fully explore known uncertainties. Here we estimate probability distributions for these projections for the SSPs using Gaussian Process emulation of the ice sheet and glacier model ensembles. We model the sea level contribution as a function of global mean surface air temperature forcing and (for the ice sheets) model parameters, with the 'nugget' allowing for multi-model structural uncertainty. Approximate independence of ice sheet and glacier models is assumed, because a given model responds very differently under different setups (such as initialisation). We find that limiting global warming to 1.5&#176;C would halve the land ice contribution to 21st century sea level rise, relative to current emissions pledges: the median decreases from 25 to 13 cm sea level equivalent (SLE) by 2100. However, the Antarctic contribution does not show a clear response to emissions scenario, due to competing processes of increasing ice loss and snowfall accumulation in a warming climate. However, under risk-averse (pessimistic) assumptions for climate and Antarctic ice sheet model selection and ice sheet model parameter values, Antarctic ice loss could be five times higher, increasing the median land ice contribution to 42 cm SLE under current policies and pledges, with the 95th percentile exceeding half a metre even under 1.5&#176;C warming. Gaussian Process emulation can therefore be a powerful tool for estimating probability density functions from multi-model ensembles and testing the sensitivity of the results to assumptions.

Download Full-text

Multidisciplinary design optimization of hard rock tunnel boring machine using collaborative optimization

Advances in Mechanical Engineering ◽

10.1177/1687814018754726 ◽

2018 ◽

Vol 10 (1) ◽

pp. 168781401875472 ◽

Cited By ~ 5

Author(s):

Wei Sun ◽

Xiaobang Wang ◽

Maolin Shi ◽

Zhuqing Wang ◽

Xueguan Song

Keyword(s):

Design Optimization ◽

Multidisciplinary Design Optimization ◽

Hard Rock ◽

Tunnel Boring Machine ◽

Collaborative Optimization ◽

Optimization Process ◽

Multidisciplinary Design ◽

Tunnel Boring ◽

Geological Conditions ◽

Rock Tunnel

A multidisciplinary design optimization model is developed in this article to optimize the performance of the hard rock tunnel boring machine using the collaborative optimization architecture. Tunnel boring machine is a complex engineering equipment with many subsystems coupled. In the established multidisciplinary design optimization process of this article, four subsystems are taken into account, which belong to different sub-disciplines/subsytems: the cutterhead system, the thrust system, the cutterhead driving system, and the economic model. The technology models of tunnel boring machine’s subsystems are build and the optimization objective of the multidisciplinary design optimization is to minimize the construction period from the system level of the hard rock tunnel boring machine. To further analyze the established multidisciplinary design optimization, the correlation between the design variables and the tunnel boring machine’s performance is also explored. Results indicate that the multidisciplinary design optimization process has significantly improved the performance of the tunnel boring machine. Based on the optimization results, another two excavating processes under different geological conditions are also optimized complementally using the collaborative optimization architecture, and the corresponding optimum performance of the hard rock tunnel boring machine, such as the cost and energy consumption, is compared and analysed. Results demonstrate that the proposed multidisciplinary design optimization method for tunnel boring machine is reliable and flexible while dealing with different geological conditions in practical engineering.

Download Full-text

Multidisciplinary Design Optimization of a Lunar Lander’s Soft-Landing Gear

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.42.118 ◽

2010 ◽

Vol 42 ◽

pp. 118-121

Author(s):

Yun Tong Lu ◽

Chun Jie Wang ◽

Ang Li ◽

Han Wang

Keyword(s):

Design Optimization ◽

Multidisciplinary Design Optimization ◽

Rapid Development ◽

Fem Analysis ◽

Landing Gear ◽

Collaborative Optimization ◽

Multidisciplinary Design ◽

Optimization Approach ◽

Soft Landing ◽

Multi Body

The rapid development of Multidisciplinary Design Optimization (MDO) approach can simultaneously guarantee the cut of cost on design and optimal performance of spacecraft. Based on the theory of Collaborative Optimization approach (CO) of MDO, present paper proposes the method of CO by integrating Pro/E(3D modeling), Patran/Nastran(FEM analysis) and ADAMS(multi-body dynamic analysis) with the Isight software. In the analysis of the soft-landing gear of Lunar Lander, this method can optimize the mass of the landing gear and meanwhile ensures the reliability of structure statics, structure dynamics and multi-body dynamics. Thus the feasibility, applied value and guideline significance of this method in spacecraft structural design are proven.

Download Full-text

Design Optimization of a Special Class of Steel Structures via Genetic Algorithms and Stochastic Sampling

Volume 5: 35th Design Automation Conference, Parts A and B ◽

10.1115/detc2009-87499 ◽

2009 ◽

Author(s):

Karim Hamza ◽

Ashraf O. Nassef ◽

Mohammed Shalaby

Keyword(s):

Genetic Algorithms ◽

Design Optimization ◽

Special Class ◽

Steel Structures ◽

Sampling Technique ◽

Manufacturing Cost ◽

Stochastic Sampling ◽

Optimization Task ◽

Consistent Performance ◽

Continuous And Discrete Variables

This paper addresses the design optimization of a special class of steel structures, which is clear-span building built up via off-shelf standard steel-sections. The problem is of particular importance in small to medium span buildings due to an attractive opportunity for reduction of the manufacturing cost compared to trusses and custom-built beams. The problem is also difficult from an optimization perspective as it exhibits both continuous and discrete variables, as well as discontinuities and flat regions in the topology of the objective function. Genetic algorithms (GA) and a special stochastic sampling technique are considered for the problem, as well as a mixed GA and stochastic sampling approach. The stochastic sampling is guided via heuristic rules based on knowledge specific to the problem, and is thus perceived well suited to the optimization task. While all the tested algorithms produced satisfactory results, the mixed approach seemed to yield the most consistent performance.

Download Full-text

Bird Species Identification Using Spectrogram Based on Multi-Channel Fusion of DCNNs

Entropy ◽

10.3390/e23111507 ◽

2021 ◽

Vol 23 (11) ◽

pp. 1507

Author(s):

Feiyu Zhang ◽

Luyang Zhang ◽

Hongxiang Chen ◽

Jiangjian Xie

Keyword(s):

Species Identification ◽

Feature Fusion ◽

Bird Species ◽

The Other ◽

Identification Accuracy ◽

Mean Average Precision ◽

Training Dataset ◽

Model Parameters ◽

Average Precision ◽

Fusion Mode

Deep convolutional neural networks (DCNNs) have achieved breakthrough performance on bird species identification using a spectrogram of bird vocalization. Aiming at the imbalance of the bird vocalization dataset, a single feature identification model (SFIM) with residual blocks and modified, weighted, cross-entropy function was proposed. To further improve the identification accuracy, two multi-channel fusion methods were built with three SFIMs. One of these fused the outputs of the feature extraction parts of three SFIMs (feature fusion mode), the other fused the outputs of the classifiers of three SFIMs (result fusion mode). The SFIMs were trained with three different kinds of spectrograms, which were calculated through short-time Fourier transform, mel-frequency cepstrum transform and chirplet transform, respectively. To overcome the shortage of the huge number of trainable model parameters, transfer learning was used in the multi-channel models. Using our own vocalization dataset as a sample set, it is found that the result fusion mode model outperforms the other proposed models, the best mean average precision (MAP) reaches 0.914. Choosing three durations of spectrograms, 100 ms, 300 ms and 500 ms for comparison, the results reveal that the 300 ms duration is the best for our own dataset. The duration is suggested to be determined based on the duration distribution of bird syllables. As for the performance with the training dataset of BirdCLEF2019, the highest classification mean average precision (cmAP) reached 0.135, which means the proposed model has certain generalization ability.

Download Full-text

Gaussian Process Regression Based Robust Optimization with Observer Uncertainty for Reconfigurable Self-x Sensory Electronics for Industry 4.0

tm - Technisches Messen ◽

10.1515/teme-2021-0061 ◽

2021 ◽

Vol 88 (s1) ◽

pp. s83-s88

Author(s):

Qummar Zaman ◽

Senan Alraho ◽

Andreas König

Keyword(s):

Robust Optimization ◽

Gaussian Process ◽

Industry 4.0 ◽

Performance Metrics ◽

Optimization Technique ◽

Gaussian Process Regression ◽

Search Space ◽

Instrumentation Amplifier ◽

Statistical Regression ◽

Current Feedback

Abstract This paper presents a robust optimization technique for the reconfigurable measurement of sensory electronics for industry 4.0 to obtain a robust solution even in the presence of observer uncertainty using a cost-effective performance measurement method. The extrinsic evaluation of the proposed methodology is performed on an indirect current-feedback instrumentation amplifier (CFIA), which is a fundamental part of sensory systems. To reduce the CFIA device performance evaluation set-up cost, a low-cost test stimulus is applied to the circuit under test, and the output response of the circuit is examined to correlate with the device’s performance parameters. Due to the complexity of the smart sensory electronics search space, the meta-heuristic optimization algorithm is being selected as an optimizer. For objective space or observer uncertainty, the Gaussian process regression from the Bayesian statistical regression process is used to estimate the uncertainty level efficiently. Six different classical metrics have been used to evaluate the regression model accuracy. The highest achieved average expected error metrics value is 0.313, and the minimum value of correlation performance metrics is 0.908. The device is implemented using 0.35 μm austriamicrosystems technology.

Download Full-text

Multi-objective collaborative multidisciplinary design optimization using particle swarm techniques and fuzzy decision making

Proceedings of the Institution of Mechanical Engineers Part C Journal of Mechanical Engineering Science ◽

10.1177/0954406211432981 ◽

2012 ◽

Vol 226 (9) ◽

pp. 2281-2295 ◽

Cited By ~ 5

Author(s):

Mohammad Reza Farmani ◽

Jafar Roshanian ◽

Meisam Babaie ◽

Parviz M Zadeh

Keyword(s):

Particle Swarm Optimization ◽

Design Optimization ◽

Multidisciplinary Design Optimization ◽

Optimization Problems ◽

Particle Swarm ◽

Collaborative Optimization ◽

Multidisciplinary Design ◽

Fuzzy Decision ◽

Swarm Optimization ◽

Multi Objective

This article focuses on the efficient multi-objective particle swarm optimization algorithm to solve multidisciplinary design optimization problems. The objective is to extend the formulation of collaborative optimization which has been widely used to solve single-objective optimization problems. To examine the proposed structure, racecar design problem is taken as an example of application for three objective functions. In addition, a fuzzy decision maker is applied to select the best solution along the pareto front based on the defined criteria. The results are compared to the traditional optimization, and collaborative optimization formulations that do not use multi-objective particle swarm optimization. It is shown that the integration of multi-objective particle swarm optimization into collaborative optimization provides an efficient framework for design and analysis of hierarchical multidisciplinary design optimization problems.

Download Full-text

Sampling behavioral model parameters for ensemble-based sensitivity analysis using Gaussian process emulation and active subspaces

Stochastic Environmental Research and Risk Assessment ◽

10.1007/s00477-020-01867-0 ◽

2020 ◽

Vol 34 (11) ◽

pp. 1813-1830

Author(s):

Daniel Erdal ◽

Sinan Xiao ◽

Wolfgang Nowak ◽

Olaf A. Cirpka

Keyword(s):

Sensitivity Analysis ◽

Gaussian Process ◽

Parameter Space ◽

Global Sensitivity Analysis ◽

Model Parameters ◽

Large Sample Size ◽

Global Sensitivity ◽

Environmental Models ◽

Additional Costs ◽

Active Subspace

Abstract Ensemble-based uncertainty quantification and global sensitivity analysis of environmental models requires generating large ensembles of parameter-sets. This can already be difficult when analyzing moderately complex models based on partial differential equations because many parameter combinations cause an implausible model behavior even though the individual parameters are within plausible ranges. In this work, we apply Gaussian Process Emulators (GPE) as surrogate models in a sampling scheme. In an active-training phase of the surrogate model, we target the behavioral boundary of the parameter space before sampling this behavioral part of the parameter space more evenly by passive sampling. Active learning increases the subsequent sampling efficiency, but its additional costs pay off only for a sufficiently large sample size. We exemplify our idea with a catchment-scale subsurface flow model with uncertain material properties, boundary conditions, and geometric descriptors of the geological structure. We then perform a global-sensitivity analysis of the resulting behavioral dataset using the active-subspace method, which requires approximating the local sensitivities of the target quantity with respect to all parameters at all sampled locations in parameter space. The Gaussian Process Emulator implicitly provides an analytical expression for this gradient, thus improving the accuracy of the active-subspace construction. When applying the GPE-based preselection, 70–90% of the samples were confirmed to be behavioral by running the full model, whereas only 0.5% of the samples were behavioral in standard Monte-Carlo sampling without preselection. The GPE method also provided local sensitivities at minimal additional costs.

Download Full-text