Near-optimal Design Research Articles

Printable nanostructured silicon solar cells for high-performance, large-area flexible photovoltaics.

Nanostructured forms of crystalline silicon represent an attractive materials building block for photovoltaics due to their potential benefits to significantly reduce the consumption of active materials, relax the requirement of materials purity for high performance, and hence achieve greatly improved levelized cost of energy. Despite successful demonstrations for their concepts over the past decade, however, the practical application of nanostructured silicon solar cells for large-scale implementation has been hampered by many existing challenges associated with the consumption of the entire wafer or expensive source materials, difficulties to precisely control materials properties and doping characteristics, or restrictions on substrate materials and scalability. Here we present a highly integrable materials platform of nanostructured silicon solar cells that can overcome these limitations. Ultrathin silicon solar microcells integrated with engineered photonic nanostructures are fabricated directly from wafer-based source materials in configurations that can lower the materials cost and can be compatible with deterministic assembly procedures to allow programmable, large-scale distribution, unlimited choices of module substrates, as well as lightweight, mechanically compliant constructions. Systematic studies on optical and electrical properties, photovoltaic performance in experiments, as well as numerical modeling elucidate important design rules for nanoscale photon management with ultrathin, nanostructured silicon solar cells and their interconnected, mechanically flexible modules, where we demonstrate 12.4% solar-to-electric energy conversion efficiency for printed ultrathin (∼ 8 μm) nanostructured silicon solar cells when configured with near-optimal designs of rear-surface nanoposts, antireflection coating, and back-surface reflector.

Heat Exchanger Network Synthesis on Gas Separation Plant No.2 (gsp2) in Thailand

Energy conservation is one of the most common concerns in gas separation plants. It plays a key role in energy and operating cost saving of their high-energy consuming processes. Although many research works on heat exchanger network (HEN) synthesis has been extensively studied for more than 40 years, most of them have limitations on computational time and feasibility due to their mathematical difficulties. Thus, they are impractical for large problems as industrial cases. In this work, a strategy for HEN synthesis was presented and applied to the industrial case of gas separation plant no.2 (GSP2) in Thailand consisting of twelve hot and eleven cold process streams. The strategy for HEN synthesis is a combination of Pinch Technology, the well-known thermodynamics-based approach, and mathematical programming based on the stage-wise superstructure model. HEN was synthesized in two parts; above and below pinch, corresponding to the optimal pinch temperature as well as the heat recovery approach temperature (HRAT) predicted by Pinch Analysis in order to assure near-optimal design. The HEN was then improved by a mathematical model applying relaxation technique based on a concept of loop and path toward better design with less complexity. According to the industrial case of GSP2, the proposed strategy for HEN synthesis is effective where a good solution can be obtained with reasonable computational time.

Using self-adaptive optimisation methods to perform sequential optimisation for low-energy building design

The use of software tools to aid building design, or to show compliance, is now commonplace. This has motivated investigations into the potential of optimisation algorithms, used in such software, to automatically optimise designs, or to generate a variety of near-optimal designs. Optimisation always requires the evaluation of a large number of possibilities, before a final selection is made. Normally when using a building simulator to assess the quality of designs, all possible solutions in the early stages of optimisation (when there is a high volume of choices) are evaluated using the same tool, so that the computational time for the assessment of each of the possibilities is the same as the time required for the final, refined choice of solutions. This paper suggests using a method of evaluation which changes as the algorithm evolves: whereas accuracy is initially compromised to improve the speed of the algorithm, the process is subsequently altered to produce a more accurate, evaluation process. This is a case of dynamic optimisation that requires an algorithm able to cope with changes in the objective landscape. A self-adaptive evolutionary strategy has been chosen, for its ability to “learn” about changes, and the influence of the different decision variables in the objective function as they arise. The results show that this method can reach the same optimal design, with substantially lower computational time than the optimisation methods found in the literature.

Emotional design method of product presented in multi-dimensional variables based on Kansei Engineering

In the fiercely competitive product market, product designs have already moved from the product-oriented approach to a marketing-oriented approach, and gradually to a customer-oriented approach. Therefore, it is of high necessity to study users’ emotional needs aroused by the product's multi-dimensional design variables. This paper contributes a systematically emotional design method of products’ hard interface based on Kansei Engineering (KE), which can be used to design a product that echoes users’ emotions. Based on the representative pairwise Kansei image words and multi-dimensional key design variables obtained by the consumer-oriented techniques, the KE models are established. Finally, the integration of KE models and Genetic Algorithm is employed to search for a near-optimal design scheme. This proposed method is effectively demonstrated in terms of the mini digital camera. It can be used in various design cases to optimise the product emotional design.

Fundamental limits and near-optimal design of graphene modulators and non-reciprocal devices

The potential of graphene for use in photonic applications was evidenced by recent demonstrations of modulators, polarisation rotators, and isolators. These promising yet preliminary results raise crucial questions: what is the optimal performance achievable by more complex designs using multilayer structures, graphene patterning, metal additions, or a combination of these approaches, and how can this optimum design be achieved in practice? Today, the complexity of the problem, which is magnified by the variability in graphene parameters, leaves the design of these new devices to time-consuming and suboptimal trial-and-error procedures. We address this issue by first demonstrating that the relevant figures of merit for the devices above are subject to absolute theoretical upper bounds. Strikingly these limits are related only to the conductivity tensor of graphene; thus, we can provide essential roadmap information such as the best possible device performance versus wavelength and graphene quality. Second, based on the theory developed, physical insight, and detailed simulations, we demonstrate how structures closely approaching these fundamental limits can be designed, demonstrating the possibility of significant improvement. These results are believed to be of paramount importance for the design of graphene-based modulators, rotators, and isolators and are also directly applicable to other 2-dimensional materials.

Using a genetic algorithm to generate D s -optimal designs for mixture experiments in a simplex region

We propose and develop a genetic algorithm(GA) to generate Ds-optimal or near-optimal mixture designs of a simplex region that maximizes the Ds-criterion while simultaneously guaranteeing an acceptable D-efficiency for the mixture experiments. Our method does not need a candidate set which makes it possible to select points throughout a continuous region. Two new GAs are developed to handle specific subsets of quadratic and special cubic mixture model terms. Summaries of GA designs are reported for 3 and 4 mixture components. The performance of GA designs is assessed in comparisons with the designs generated from an exchange algorithm (EA) and SAS Proc OPTEX computer generated designs (CGDs). The results show that the Ds-criterion values of the GA designs for all percentage of D-efficiency were greater than or equal to those of the EA and SAS Proc OPTEX designs(CGDs). This suggests that the GA is a effective method for generating subset optimal designs for mixture experiments.

A new laminates encoding scheme for the genetic algorithm-based optimization of stiffened composite panels

Purpose – The genetic algorithm (GA) technique is widely used for the optimization of stiffened composite panels. It is based on sequential execution of a number of specific operators, including the encoding of particular design variables. For instance, in the case of a stiffened composite panel, the design variables that need to be encoded are: the number of plies and their stacking sequences in the panel skin and stiffeners. This paper aims to present a novel, implicit, heuristic approach for encoding composite laminates and, through its use, demonstrates an improvement in the optimization process. Design/methodology/approach – The stiffened panel optimization has been formulated as a constrained discrete minimum-weight design problem. GAs, which use both new encoding schemes and those previously described in the literature, have been used to find near-optimal solutions to the formulated problem. The influence of the new encoding scheme on the searching capabilities of the GA has been investigated using comparative analysis of the optimization results. Findings – The new encoding scheme allows the definition of stacking sequences in composites using shorter symbolic representations as compared with standard encoding operators and, as a result of this, a reduction in the problem design space. According to numerical experiments performed in this work, this feature enables GA to obtain near-optimal designs using smaller population sizes than those required if standard encoding schemes are used. Originality/value – The approach to encoding laminates presented in this paper is based on the original heuristics. In the context of GA-based optimization of stiffened composite panels, the use of the new approach rather than the standard encoding technique can lead to a significant reduction in computational time employed.

Conceptual design of heterogeneous azeotropic distillation process for ethanol dehydration using 1-butanol as entrainer

The synthesis of a heterogeneous azeotropic distillation process for ethanol dehydration using 1-butanol as entrainer is presented. The residue curve map of the ethanol/water/1-butanol mixture is computationally generated using non-random two- liquid thermodynamic model. It is found that 1-butanol leads to a residue curve map topological structure different from that generated by typical entrainers used in ethanol dehydration. Synthesised by residue curve map analysis, the distillation flowsheet for ethanol dehydration by 1-butanol comprises a double-feed column integrated with an overhead decanter and a simple column. The double-feed column is used to recover water as the top product, whereas the simple column is used for recovering ethanol and 1- butanol. The separation feasibility and the economically near-optimal designs of distillation columns in the flowsheet are evaluated and identified by using the boundary value design method. The distillation flowsheet using 1-butanol is compared with the conventional process using benzene as entrainer. Based on their total annualised costs, the ethanol dehydration process using 1-butanol is less economically attractive than the process using benzene. However, 1-butanol is less toxic than benzene.

Optimization of Hierarchical Modulation for Decode-and-Forward Wireless Relay Networks

This paper presents two designs of optimal nonuniformed constellations for decode-and-forward wireless relay networks with an orthogonal space-time block code (STBC). The first design is concerned with the unequal error protection issue, in which two data streams to be transmitted from a source to a destination are protected at two different levels. The design is to minimize the bit error rate (BER) of one stream while maintaining the BER of the other stream to be no larger than a given threshold. Based on a simple approximation of the average BER, a design parameter is obtained in closed form. The second design focuses on minimizing the average BER of the combined data stream at the destination. Developed are two near-optimal designs for 4/16-quadrature amplitude modulation (QAM) and 4/64-QAM hierarchical modulation (HM) schemes. The designs are also extended to two-way relay networks where both the source and the destination have data to transmit to each other. Simulation results confirm the superiority of the proposed designs and the advantages of relay-assisted transmission employing HM over conventional point-to-point transmission. A comparison with the existing relay-assisted transmission model is also discussed.

Combining liquid desiccant dehumidification with a dew-point evaporative cooler: A design analysis

This article uses a numerical model to analyze a concept combining a liquid desiccant dehumidifier with a dew-point indirect evaporative cooler. Each of these components, or stages, consists of an array of channel pairs, where a channel pair is two air channels separated by a thin plastic plate. In the first stage, a liquid desiccant film lining one side of the plates removes moisture from the process (supply-side) air through a membrane. An evaporatively cooled exhaust airstream on the other side of the plastic plate cools the desiccant. The second stage sensibly cools the dried process air with a dew-point evaporative cooler. This article uses a parametric analysis to illustrate the key design tradeoff for this concept: device size (a surrogate for cost) versus energy efficiency. The analysis finds the design parameters with the largest effect on this tradeoff and finds the combinations of design parameters giving near-optimal designs, which are designs with the highest efficiency for a given device size. The results indicate that there are two key parameters contributing to this tradeoff: the supply-side air channel thickness and the exhaust-air flow rate in the evaporative cooler.

Near-optimal experimental design for model selection in systems biology.

Motivation: Biological systems are understood through iterations of modeling and experimentation. Not all experiments, however, are equally valuable for predictive modeling. This study introduces an efficient method for experimental design aimed at selecting dynamical models from data. Motivated by biological applications, the method enables the design of crucial experiments: it determines a highly informative selection of measurement readouts and time points.Results: We demonstrate formal guarantees of design efficiency on the basis of previous results. By reducing our task to the setting of graphical models, we prove that the method finds a near-optimal design selection with a polynomial number of evaluations. Moreover, the method exhibits the best polynomial-complexity constant approximation factor, unless P = NP. We measure the performance of the method in comparison with established alternatives, such as ensemble non-centrality, on example models of different complexity. Efficient design accelerates the loop between modeling and experimentation: it enables the inference of complex mechanisms, such as those controlling central metabolic operation.Availability: Toolbox ‘NearOED’ available with source code under GPL on the Machine Learning Open Source Software Web site (mloss.org).Contact: busettoa@inf.ethz.chSupplementary information: Supplementary data are available at Bioinformatics online.

Examining the Possibilities: Generating Alternative Watershed-Scale BMP Designs with Evolutionary Algorithms

Recent studies in water resources planning and management show a gradual shift in the state of the art from numerous on-site structural stormwater Best Management Practice (BMP) designs to watershed-scale BMP design approaches that meet both target water quantity (peak flow) and quality (sediment reduction) criteria. Such regionally-strategic approaches are not only cost-effective but emphasize comprehensive, holistic watershed-scale management, an idea strongly promoted by the U.S. Environmental Protection Agency (US EPA) since the early 1990s. Implementing a watershed-scale design can prove difficult when decision-makers have differing and sometimes conflicting objectives. We present a methodology that integrates the semi-distributed watershed model Soil and Water Assessment Tool (SWAT) with an evolutionary algorithm, Species Conserving Genetic Algorithm (SCGA). In addition to identifying an optimal watershed-scale BMP design (e.g., type, size, location), SCGA simultaneously produces several near-optimal design alternatives using a user-specified distance metric. We demonstrate this decision-oriented framework on a watershed in southern Illinois. Results of this application yield several high-quality alternative designs appropriate for solving integrated watershed management problems.

Pareto simulated annealing for the design of experiments: illustrated by a gene expression study

Abstract Experimental design is concerned with the problem of allocating resources within an experiment to ensure that objectives of the experiment are achieved at the minimum cost. This paper focuses on the generation of optimal or near-optimal designs for large and complex experiments where it is infeasible to carry out an ex- haustive search of the design space. Optimal designs for gene expression studies, aimed at investigating the behaviour of genes, are considered, where the optimality criterion employed is Pareto optimality. We develop an adaptation of the metaheuris- tic method of Pareto simulated annealing to generate an approximation to the set of Pareto optimal designs for large and complex experiments. We develop algorithms that utilise response surface methodology to search systematically for the optimal values of parameters associated with Pareto simulated annealing and performance is evaluated using quality measures.

Empirical Pseudo-Optimal Waveform Design for Dispersive Propagation Through Loamy Soil

Using a fast Fourier transform-based theoretical formulation as a starting point for a near-optimal waveform design, three near-optimal or pseudo-optimal pulses are empirically derived in this letter to achieve near-optimal penetration depth through the sample of loamy soil considered in this study. We describe three methods to derive these waveforms with improved features valid for any dispersive media: the first of them establishes a close approach of the Brillouin pulses; and the other two analyze structures that reinforce the Brillouin precursor formation to simultaneously achieve improved peak decay and reduced pulse broadening. The pulses are tested for experimental data achieved within a reduced range of frequency, 0.5-3 GHz, even that any theoretical dielectric model can be also used.

적응형 유전알고리즘을 이용한 사용자 인터페이스 설계 방법 개발

The size and layout of user interface components need to be optimally designed in terms of reachability, visibility, clearance, and compatibility in order for efficient and effective use of products. The present study develops an ergonomic design method which optimizes the size and layout of user interface components using adaptive genetic algorithm. The developed design method determines a near-optimal design which maximizes the aggregated score of 4 ergonomic design criteria (reachability, visibility, clearance, and compatibility). The adaptive genetic algorithm used in the present study finds a near-optimum by automatically adjusting the key parameter (probability of mutation) of traditional genetic algorithm according to the characteristic of current solutions. Since the adaptive mechanism partially helps to overcome the local optimality problem, the probability of finding the near-optimum has been substantially improved. To evaluate the effectiveness of the developed design method, the present study applied it to the user interface design for a portable wireless communication radio.

Optimal design for multi-arm multi-stage clinical trials

In early stages of drug development there is often uncertainty about the most promising among a set of different treatments. In order to ensure the best use of resources it is important to decide which, if any, of the treatments should be taken forward for further testing. Multi-arm multi-stage (MAMS) trials provide gains in efficiency over separate randomised trials of each treatment. They allow a shared control group, dropping of ineffective treatments before the end of the trial and stopping the trial early if sufficient evidence of a treatment being superior to control is found. Moreover a direct, head-to-head, comparison of treatments is undertaken that ensures that many potentially influential outside factors are eliminated. In this talk we discuss optimal MAMS designs for normally distributed endpoints. An optimal design has the required type I error rate and power, but minimises the expected sample size (ESS) at some combination of treatment effects. Finding an optimal design requires searching over the stopping boundaries and sample size per stage, potentially a large number of parameters. We propose a method which combines quick evaluation of specific designs and an efficient stochastic search for the optimal design parameters. The search can also take the allocation ratio between controls and active treatments into account, allowing further efficiency gains. In the two-arm case, the triangular design has excellent expected sample size properties, and is immediate to find. Here we find that there is potential for greater improvements over the triangular design, especially as the number of stages or active treatments increases. The triangular design still serves as a quick-to-find and near-optimal design, so may still be useful for design of MAMS trials.

Design Optimization of Vehicle Control Networks

The advancement of electronic technology has made significant contributions to the safety and convenience of modern vehicles. New intelligent functionalities of vehicles have been implemented in a number of electronic control units (ECUs) that are connected to each in vehicle control networks (VCNs). However, with the rapid increase in the number of ECUs, VCNs currently face several challenges, e.g., design complexity, space constraints, system reliability, and interdependency. Considering these factors, the complexity of the VCN design problem exponentially increases, which means that the problem cannot be solved within a reasonable time using conventional optimization techniques. In this paper, we report a new methodology for the optimal design of VCNs. An analytical model was derived to examine the fundamental characteristics of the VCN design problem. Compared with the case of a conventional data network, which typically considers temporal scheduling over a fixed physical topology, the VCN design problem should also consider spatial constraints, e.g., volume, position, and weight. Moreover, the spatial constraints change during the solving procedure. Such temporal and spatial joint optimization problems with varying constraints incur extremely high computational complexity. To tackle the high complexity, this paper proposes a fast solution based on a repeated-matching method, which reduces the problem complexity from <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">O</i> ( <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">NNN</i> ) to <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">O</i> ( <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">NN</i> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> ). By applying our methodology to a number of different real-world VCN design scenarios, this proposal can produce a 1% near-optimal design within a significantly reduced time.

Detector Design and Complexity Analysis for Cooperative Communications in Intersymbol Interference Fading Channels

The area of cooperative communications has recently attracted lots of research interest because of the potential benefit of increased spatial diversity. In this paper, we study the detector design for cooperative communications in intersymbol interference (ISI) channels. A novel system framework employing nonorthogonal amplify-and-forward half-duplex relays through ISI channels is introduced. We focus on detector design and first consider an optimal detector that consists of a whitening filter and a maximum-likelihood sequence estimator (MLSE). However, such an optimal detector embedded with Viterbi algorithm has practical issues of high complexity if the relay period is long. Consequently, we adopt a multitrellis Viterbi algorithm (MVA) that reduces the complexity significantly but still achieves near-optimal performance. Simulation results demonstrate the performance of both the optimal and near-optimal detector designs.

Practical considerations for experimental designs of spatially autocorrelated data using computer intensive methods

Classical experimental design depends upon randomization of treatment applications. However, if data come from a spatially autocorrelated random process, it is possible to find specific designs that are much better. For example, it is possible to find universally optimum complete block designs that are optimal for a spatial process where the variables are independent among blocks but within blocks follow an autoregressive second order process. The problem with these designs is that they are only possible for certain combinations of plots, blocks, and treatments, and the optimality criteria is based on a particular set of contrasts. In this paper, I use methods based on simple genetic algorithms with simulating annealing to find good experimental designs for any combination of plots, blocks, treatments, and any set of contrasts. The computer intensive methods find optimal designs equivalent to universally optimum complete block design for certain sets of contrasts, and they find better designs for other sets of contrasts. The computer intensive methods are much better than randomized designs. I show a real example for a 2×3 factorial experiment where shade and water treatments were applied to examine their effects on caribou forage quality, where I also examine the robustness of the design. General guidelines are discussed on using computer intensive methods to find near-optimal designs.

Design and Analysis of High-Index-Contrast Gratings Using Coupled Mode Theory

We analyze high-refractive-index-contrast subwavelength grating structures using truncated coupled mode theory (CMT). CMT not only provides physical insight into the role of each mode in the overall response but also allows for improved design. An analytic expression is derived for the design of broadband reflectors, providing a near-optimal design that is within 0.08% of the maximum broadband reflectivity calculated by the finite-difference time-domain method. Furthermore, the CMT is used to design a high-quality narrow-band reflector with 28% improved quality factor over previously reported results, as quantified by rigorous coupled wave analysis.