Stochastic Optimization Procedure Research Articles

Accurate stochastic reconstruction of heterogeneous microstructures by limited x‐ray tomographic projections

An accurate knowledge of the complex microstructure of a heterogeneous material is crucial for its performance prediction, prognosis and optimization. X-ray tomography has provided a nondestructive means for microstructure characterization in 3D and 4D (i.e. structural evolution over time), in which a material is typically reconstructed from a large number of tomographic projections using filtered-back-projection (FBP) method or algebraic reconstruction techniques (ART). Here, we present in detail a stochastic optimization procedure that enables one to accurately reconstruct material microstructure from a small number of absorption contrast x-ray tomographic projections. This discrete tomography reconstruction procedure is in contrast to the commonly used FBP and ART, which usually requires thousands of projections for accurate microstructure rendition. The utility of our stochastic procedure is first demonstrated by reconstructing a wide class of two-phase heterogeneous materials including sandstone and hard-particle packing from simulated limited-angle projections in both cone-beam and parallel beam projection geometry. It is then applied to reconstruct tailored Sn-sphere-clay-matrix systems from limited-angle cone-beam data obtained via a lab-scale tomography facility at Arizona State University and parallel-beam synchrotron data obtained at Advanced Photon Source, Argonne National Laboratory. In addition, we examine the information content of tomography data by successively incorporating larger number of projections and quantifying the accuracy of the reconstructions. We show that only a small number of projections (e.g. 20-40, depending on the complexity of the microstructure of interest and desired resolution) are necessary for accurate material reconstructions via our stochastic procedure, which indicates its high efficiency in using limited structural information. The ramifications of the stochastic reconstruction procedure in 4D materials science are also discussed.

A stochastic optimization framework for the restoration of an over-exploited aquifer

ABSTRACTThis study investigates the impact of hydraulic conductivity uncertainty on the sustainable management of the aquifer of Lake Karla, Greece, using the stochastic optimization approach. The lack of surface water resources in combination with the sharp increase in irrigation needs in the basin over the last 30 years have led to an unprecedented degradation of the aquifer. In addition, the lack of data regarding hydraulic conductivity in a heterogeneous aquifer leads to hydrogeologic uncertainty. This uncertainty has to be taken into consideration when developing the optimization procedure in order to achieve the aquifer’s sustainable management. Multiple Monte Carlo realizations of this spatially-distributed parameter are generated and groundwater flow is simulated for each one of them. The main goal of the sustainable management of the ‘depleted’ aquifer of Lake Karla is two-fold: to determine the optimum volume of renewable groundwater that can be extracted, while, at the same time, restoring its water table to a historic high level. A stochastic optimization problem is therefore formulated, based on the application of the optimization method for each of the aquifer’s multiple stochastic realizations in a future period. In order to carry out this stochastic optimization procedure, a modelling system consisting of a series of interlinked models was developed. The results show that the proposed stochastic optimization framework can be a very useful tool for estimating the impact of hydraulic conductivity uncertainty on the management strategies of a depleted aquifer restoration. They also prove that the optimization process is affected more by hydraulic conductivity uncertainty than the simulation process.Editor Z.W. Kundzewicz; Guest editor S. Weijs

Flood Catastrophe Model for Designing Optimal Flood Insurance Program: Estimating Location-Specific Premiums in the Netherlands.

As flood risks grow worldwide, a well-designed insurance program engaging various stakeholders becomes a vital instrument in flood risk management. The main challenge concerns the applicability of standard approaches for calculating insurance premiums of rare catastrophic losses. This article focuses on the design of a flood-loss-sharing program involving private insurance based on location-specific exposures. The analysis is guided by a developed integrated catastrophe risk management (ICRM) model consisting of a GIS-based flood model and a stochastic optimization procedure with respect to location-specific risk exposures. To achieve the stability and robustness of the program towards floods with various recurrences, the ICRM uses stochastic optimization procedure, which relies on quantile-related risk functions of a systemic insolvency involving overpayments and underpayments of the stakeholders. Two alternative ways of calculating insurance premiums are compared: the robust derived with the ICRM and the traditional average annual loss approach. The applicability of the proposed model is illustrated in a case study of a Rotterdam area outside the main flood protection system in the Netherlands. Our numerical experiments demonstrate essential advantages of the robust premiums, namely, that they: (1) guarantee the program's solvency under all relevant flood scenarios rather than one average event; (2) establish a tradeoff between the security of the program and the welfare of locations; and (3) decrease the need for other risk transfer and risk reduction measures.

Asymptotic Behavior of a Modified Stochastic Optimization Procedure in an Averaging Scheme

The asymptotic behavior of a modified discrete stochastic optimization procedure (SOP) in a Markov environment in an averaging scheme is investigated. Additional SOP optimization parameters are introduced and are used to investigate the behavior of fluctuations on increasing time intervals. The normalization-dependent form of the limit generator is established, and a modified discrete SOP is shown to be asymptotically normal when the introduced parameters assume certain values.

Robust optimum design of base isolation system in seismic vibration control of structures under random system parameters

The optimum design of base isolation system to control seismic vibration considering uncertain system parameters are usually performed by minimizing the unconditional expected value of mean square response of a structure without any consideration to the variance of such responses due to system parameter uncertainty. However, the unconditional mean square response based designed may have larger variance of responses due to uncertainty in system parameters and the overall system performance may be sensitive. But, it is desirable that the optimum design should reduce both the mean and variance of dynamic performance measure under system parameter uncertainty. The present study deals with robust design optimization (RDO) of base isolation system considering random system parameters characterizing the structure, isolator and ground motion model. The RDO is performed by minimizing the weighted sum of the expected value of the maximum root mean square acceleration of the structure as well its standard deviation. A numerical study elucidates the importance of the RDO procedure for design of base isolation system by comparing the proposed RDO results with the results obtained by the conventional stochastic structural optimization procedure and the unconditional response based optimization.

Robust design of liquid column vibration absorber in seismic vibration mitigation considering random system parameter

The optimum design of liquid column dampers in seismic vibration control considering system parameter uncertainty is usually performed by minimizing the unconditional response of a structure without any consideration to the variation of damper performance due to uncertainty. However, the system so designed may be sensitive to the variations of input system parameters due to uncertainty. The present study is concerned with robust design optimization (RDO) of liquid column vibration absorber (LCVA) considering random system parameters characterizing the primary structure and ground motion model. The RDO is obtained by minimizing the weighted sum of the mean value of the root mean square displacement of the primary structure as well as its standard deviation. A numerical study elucidates the importance of the RDO procedure for design of LCVA system by comparing the RDO results with the results obtained by the conventional stochastic structural optimization procedure and the unconditional response based optimization.

Quantum-Inspired Evolutionary Algorithm for Topology Optimization of Modular Cabled-Trusses

This article proposes an alternative optimization framework applied to topology optimization of modular lightweight cabled-truss structures. These structures are described as a system of intrinsically positioned cables and triangular bar formations jointed at their ends by hinged connections to form a rigid framework. The optimized topologies are determined through a stochastic discrete optimization procedure that uses ground structure approach, nonlinear finite element analysis, and quantum-inspired evolutionary algorithms. The optimization searches for optimal mass reduction with minimal losses in stiffness, such relation, is expressed by the stiffness-to-mass ratio parameter. Nonlinear finite element analysis is used to evaluate the static structural response. In order to decrease computation time, kinematically instable and structurally invalid individuals are filtered before evaluation. Modular design approach is taken into consideration to reduce the number of design variables and increase the productibility of cabled-trusses. Symmetric structural response is desired since in several mechanical applications forces can assume different directions during the working cycle. A modular ground structure with 300 elements is optimized, and optimal truss and cabled-truss topologies are compared. Complementary analyses comprise the investigation of the structural performance under different number of modules and slenderness ratios. The results indicate that the proposed optimization framework leads to optimized structures. In addition, it was observed that cabled trusses presented significant improvements in structural performance when compared with trusses.

Inference for dynamic and latent variable models via iterated, perturbed Bayes maps

Iterated filtering algorithms are stochastic optimization procedures for latent variable models that recursively combine parameter perturbations with latent variable reconstruction. Previously, theoretical support for these algorithms has been based on the use of conditional moments of perturbed parameters to approximate derivatives of the log likelihood function. Here, a theoretical approach is introduced based on the convergence of an iterated Bayes map. An algorithm supported by this theory displays substantial numerical improvement on the computational challenge of inferring parameters of a partially observed Markov process.

Design, production and reverse engineering of ultra-steep hot mirrors.

We present the whole design-production chain of an ultra-steep hot mirror produced using the indirect monochromatic monitoring technique. The hot mirror without thin layers is designed utilizing the stochastic optimization procedure that takes in account upper and lower constraints for layer optical thickness. We produced the hot mirror with the ion-assisted electron beam deposition technique using indirect monochromatic monitoring strategy, performed reverse engineering of the deposited coatings, and illustrated that the random variation of the tooling factors in low-index layers is the main factor causing production errors. We modified the monitoring strategy with low-index layers monitored by quartz crystal monitor, and demonstrated the excellent correspondence to the theoretical spectral performance.

A Developmental Approach to Predicting Neuronal Connectivity from Small Biological Datasets: A Gradient-Based Neuron Growth Model

Relating structure and function of neuronal circuits is a challenging problem. It requires demonstrating how dynamical patterns of spiking activity lead to functions like cognitive behaviour and identifying the neurons and connections that lead to appropriate activity of a circuit. We apply a “developmental approach” to define the connectome of a simple nervous system, where connections between neurons are not prescribed but appear as a result of neuron growth. A gradient based mathematical model of two-dimensional axon growth from rows of undifferentiated neurons is derived for the different types of neurons in the brainstem and spinal cord of young tadpoles of the frog Xenopus. Model parameters define a two-dimensional CNS growth environment with three gradient cues and the specific responsiveness of the axons of each neuron type to these cues. The model is described by a nonlinear system of three difference equations; it includes a random variable, and takes specific neuron characteristics into account. Anatomical measurements are first used to position cell bodies in rows and define axon origins. Then a generalization procedure allows information on the axons of individual neurons from small anatomical datasets to be used to generate larger artificial datasets. To specify parameters in the axon growth model we use a stochastic optimization procedure, derive a cost function and find the optimal parameters for each type of neuron. Our biologically realistic model of axon growth starts from axon outgrowth from the cell body and generates multiple axons for each different neuron type with statistical properties matching those of real axons. We illustrate how the axon growth model works for neurons with axons which grow to the same and the opposite side of the CNS. We then show how, by adding a simple specification for dendrite morphology, our model “developmental approach” allows us to generate biologically-realistic connectomes.

Continuous Stochastic Optimization with Semi-Markov Switchings in the Diffusion Approximation Scheme

We consider the continuous stochastic optimization procedure with semi-Markov switching in the diffusion approximation scheme with the balance conditions imposed on singular perturbation of the regression function. The sufficient convergence conditions are established for the regression function, which depends on the uniform ergodic semi-Markov process, by using the properties of extended compensating operator of the Markov renewal of the procedure and its asymptotic representation of perturbed Lyapunov function.

Design, optimization and controllability of an alternative process based on extractive distillation for an ethane–carbon dioxide mixture

Alternative configurations based on cryogenic extractive distillation were proposed and simulated by using Aspen Plus 7.0® coupled to a multi-objective stochastic optimization procedure (differential evolution, DE). The evaluation of the performances of the proposed configurations was focused on the ethane–carbon dioxide azeotrope separation considering different liquefied hydrocarbon fractions as entrainers. The design alternatives were compared to the conventional chemical absorption system.The proposed sequences were simultaneously Pareto optimized by minimizing the total annual cost (TAC) and maximizing the acid gas removal. Complementary studies regarding the theoretical control properties, the thermodynamic efficiency and the greenhouse gases generation were conducted for several representative operating conditions obtained from the Pareto optimized fronts. The proposed cryogenic extractive distillation sequences realized the higher carbon dioxide removal together with the lower TAC compared to the conventional chemical absorption system.

A difference stochastic optimization procedure with impulse perturbation

The sufficient conditions are obtained for the convergence of the difference stochastic optimization procedure with impulse perturbations in a Markov environment under the conditions of exponential stability of the averaged system and smooth regression function of the source system. To this end, an asymptotic representation of the perturbed procedure generator is obtained.

Fitting of self-exciting threshold autoregressive moving average nonlinear time-series model through genetic algorithm and development of out-of-sample forecasts

Self-exciting threshold autoregressive moving average (SETARMA) nonlinear time-series model is considered here. Sufficient conditions for invertibility and stationarity are derived. Parameter estimation algorithm is developed by employing real-coded genetic algorithm stochastic optimization procedure. A significant feature of the work done is that optimal out-of-sample forecasts up to three-step ahead and their forecast error variances are derived analytically. Relevant computer programs are written in statistical analysis system (SAS) and C. As an illustration, annual mackerel catch time-series data are considered. Forecast performance of the fitted model for hold-out data is evaluated by using Naive and Monte Carlo approaches. It is found that optimal out-of-sample forecast values are quite close to actual values and estimated variances are quite close to theoretical values. Superiority of the SETARMA model over the SETAR model for equal predictive ability through Diebold–Mariano test is also established.

Exhaustively characterizing feasible logic models of a signaling network using Answer Set Programming.

Motivation: Logic modeling is a useful tool to study signal transduction across multiple pathways. Logic models can be generated by training a network containing the prior knowledge to phospho-proteomics data. The training can be performed using stochastic optimization procedures, but these are unable to guarantee a global optima or to report the complete family of feasible models. This, however, is essential to provide precise insight in the mechanisms underlaying signal transduction and generate reliable predictions.Results: We propose the use of Answer Set Programming to explore exhaustively the space of feasible logic models. Toward this end, we have developed caspo, an open-source Python package that provides a powerful platform to learn and characterize logic models by leveraging the rich modeling language and solving technologies of Answer Set Programming. We illustrate the usefulness of caspo by revisiting a model of pro-growth and inflammatory pathways in liver cells. We show that, if experimental error is taken into account, there are thousands (11 700) of models compatible with the data. Despite the large number, we can extract structural features from the models, such as links that are always (or never) present or modules that appear in a mutual exclusive fashion. To further characterize this family of models, we investigate the input–output behavior of the models. We find 91 behaviors across the 11 700 models and we suggest new experiments to discriminate among them. Our results underscore the importance of characterizing in a global and exhaustive manner the family of feasible models, with important implications for experimental design.Availability: caspo is freely available for download (license GPLv3) and as a web service at http://caspo.genouest.org/.Supplementary information: Supplementary materials are available at Bioinformatics online.Contact: santiago.videla@irisa.fr

An Adaptive Sequential Monte Carlo Sampler

Sequential Monte Carlo (SMC) methods are not only a popular tool in the analysis of state–space models, but offer an alternative to Markov chain Monte Carlo (MCMC) in situations where Bayesian inference must proceed via simulation. This paper introduces a new SMC method that uses adaptive MCMC kernels for particle dynamics. The proposed algorithm features an online stochastic optimization procedure to select the best MCMC kernel and simultaneously learn optimal tuning parameters. Theoretical results are presented that justify the approach and give guidance on how it should be implemented. Empirical results, based on analysing data from mixture models, show that the new adaptive SMC algorithm (ASMC) can both choose the best MCMC kernel, and learn an appropriate scaling for it. ASMC with a choice between kernels outperformed the adaptive MCMC algorithm of Haario et al. (1998) in 5 out of the 6 cases considered.

Modern design tools and a new paradigm in optical coating design

Several modern optical coating designs tools are discussed in the frame of a new design paradigm proposing the search not for a formally optimal solution with the lowest possible merit function value but for the most practical solution that takes into account additional feasibility demands. Considered design tools include a stochastic optimization procedure that takes into account upper and lower constraints for layer optical thicknesses. This procedure allows one to obtain multiple solutions to a design problem, which presents additional opportunities for choosing a practically optimal design. Two special design techniques involving integer optimization also take into account additional demands. The first one is aimed at designing multicavity narrow bandpass filters with quarter wave or multiple quarter wave layer optical thicknesses. It enables obtaining bandpass filters with extremely steep transmittance slopes, bandwidths of several tens of nanometers, and very small ripples in transmission zones. The second technique is aimed at covering design problems that have been traditionally solved using the theory of equivalent layers. One more technique considered in this paper is aimed at reducing the influence of noncorrelated thickness errors on design spectral characteristics.

A Stochastic Optimization Procedure Applied to Ferroelectrics for Piezoelectric Applications

A stochastic optimization procedure incorporating a continuum modelling is used to identify the optimal orientation distribution of ferroelectrics (FEs) for piezoelectric applications. The orientation of ferroelectric crystals plays a critical role in the anisotropy of their piezoelectric properties. Crystallographic orientation is inextricably related to the piezoelectric properties of FEs and is characterised through Euler angles (φ, θ, ψ). Ferroelectric single crystal in general exhibit orientation dependent piezoelectricity. The macroscopic properties of a ceramic FE, in general, differ significantly from those of single crystals mainly due to the imperfect alignment of the crystallographic axes of the constituent domains or crystallites. This suggests that piezoelectric properties can be tailored by a proper choice of the parameters which control the orientation distribution. Nevertheless, this choice is complicated and it is impossible to analyze all possible combinations of the distribution parameters or the angles themselves. The set of combination of variables, known as solution space, which dictates the orientation distribution of crystallites is unlimited. Stochastic optimization combined with a generalized Monte Carlo scheme optimizes the objective functions, the effective piezoelectric coefficients d jν and the electromechanical coupling. These objective functions are calculated using the homogenization method at each orientation configuration chosen by the optimization algorithm. A modified simulated annealing is employed to optimize the objective functions described in the optimization. Polycrystalline ferroelectric materials are shown to have the potential of exhibiting better performance at a macroscopic scale by the design of the grain configuration at the micro scale.

Randomized Smoothing for Stochastic Optimization

We analyze convergence rates of stochastic optimization algorithms for nonsmooth convex optimization problems. By combining randomized smoothing techniques with accelerated gradient methods, we obtain convergence rates of stochastic optimization procedures, both in expectation and with high probability, that have optimal dependence on the variance of the gradient estimates. To the best of our knowledge, these are the first variance-based rates for nonsmooth optimization. We give several applications of our results to statistical estimation problems and provide experimental results that demonstrate the effectiveness of the proposed algorithms. We also describe how a combination of our algorithm with recent work on decentralized optimization yields a distributed stochastic optimization algorithm that is order-optimal.

Optimización robusta en aplicaciones aeronáuticas con la combinación de cálculo estocástico y algoritmos evolutivos

Uncertainties are a daily issue to deal with in aerospace engineering and applications. Robust optimization methods commonly use a random generation of the inputs and take advantage of multi-point criteria to look for robust solutions accounting with uncertainty definition. From the computational point of view, the application to coupled problems, like fluid-dynamics (CFD) or fluid-structure interaction (FSI), can be extremely expensive. This work presents a coupling between stochastic analysis techniques and evolutionary optimization algorithms for the definition of a stochastic robust optimization procedure. At first, a stochastic procedure is proposed to be applied into optimization problems. The proposed method has been applied to both CFD and FSI problems for the reduction of drag and flutter, respectively.