Numerical Optimization Techniques Research Articles

Efficient CZTSSe thin film solar cell employing MoTe2/MoS2 as hole transport layer

Substantial and perpetual growth of solar cell needs plentiful non-toxic materials such as silicon. CZTSSe material brings about new age solar cell which is alluring participant for high performance in comparison to other semiconductor materials. Since CZTSSe is non-toxic in nature and abundant on the earth surface and cheaper in cost, it has acceptable bandage reflecting eminent cell performance in the era of renewable source of energy. Recorded efficacy of such type of solar cell has festered up to 12% over many years. Non-radiative recombination at the heterojunction interface is the possible reason for the reduced output performance and lower output voltage. A numerical optimization technique using SCAPS-1D program has been used to evaluate the overall performance of the cell. In present work, three type of cell configuration have been studied they are cell without hole transport layer and cell with hole transport layer of MoTe2 and MoS2. Thickness of absorber layer, interfacial layer, for the solar cell ZnO/CdS/CZTSSe//(MoS2 or MoTe2)/Ni/glass has been investigated. In the present study, MoS2 proves itself to be a better candidate for hole transport layer for fabricating CZTSSe solar cell. Operating temperature, work function and series and shunt resistance and defect states and acceptor concentration have been optimized with the yield of 21.12% efficiency. The experimental work might be assisted to the proposed work and the findings in this work might be a great interest to experimentalists to this area.

Optimization of flour-replacing ingredients for low-carbohydrate, gluten-free muffins via a mixture design with complete sucrose substitution by d-allulose or d-tagatose

Consumer acceptance testing found wheat flour alternatives can be combined with rare sugars to create acceptable lower sugar, gluten-free muffins. Muffins were developed and optimized using an initial screening procedure. A simplex centroid mixture design defined combinations of wheat-flour alternates. Ninety-one randomized baked blends were analyzed and compared for physical and basic sensory attributes using numerical optimization techniques. Blends of almond flour, whey protein concentrate, and a 50/50 blend of oat fibre and resistant maltodextrin achieved desired structures for high-protein and high-fibre muffins. Combinations of non-dairy coconut creamer, whey protein concentrate, and the same blend of oat fibre and maltodextrin composed the high-fat variant. Optimized flour replacement blends enabled a sugar replacement comparison in a separate categorical design. Vanilla-flavored muffins formulated with monosaccharides (allulose or tagatose) or sucrose (as a control), had sweetness levels augmented with stevia to theoretically equivalent levels. Muffins with allulose or tagatose differed from those containing sucrose in water activity and crust colour (lightness, L*). Overall liking for sucrose substituted formulations compared favorably in a consumer acceptance panel (n = 58). After providing the nutritional composition for sampled products at the conclusion of the sensory analysis, purchase intent of sugar substituted products was comparable to the control muffins.

Technical note: Graph-theory-based heuristics to aid in the implementation of optimized drinking water network sectorization

Abstract. Drinking water distribution networks form an essential part of modern-day critical infrastructure. Sectorizing a network into district metered areas is a key technique for pressure management and water loss reduction. Sectorizing an existing network from scratch is, however, an exceedingly complex design task that designs in a well-studied general mathematical problem. Numerical optimization techniques such as evolutionary algorithms can be used to search for near-optimal solutions to such problems, but doing so within a reasonable time frame remains an ongoing challenge. In this work, we introduce two heuristic tricks that use information of the network structure and information of the operational requirements of the drinking water distribution network to modify the basic evolutionary algorithm used to solve the general problem. These techniques not only reduce the time required to find good solutions but also ensure that these solutions better match the requirements of drinking water practice. Both techniques were demonstrated by applying them in the sectorization of the actual distribution network of a large city.

Parameter identification of a delayed infinite-dimensional heat-exchanger process based on relay feedback and root loci analysis

The focus of this contribution is twofold. The first part aims at the rigorous and complete analysis of pole loci of a simple delayed model, the characteristic function of which is represented by a quasi-polynomial with a non-delay and a delay parameter. The derived spectrum constitutes an infinite set, making it a suitable and simple-enough representative of even high-order process dynamics. The second part intends to apply the simple infinite-dimensional model for relay-based parameter identification of a more complex model of a heating–cooling process with heat exchangers. Processes of this type and construction are widely used in industry. The identification procedure has two substantial steps. The first one adopts the simple model with a low computational effort using the saturated relay that provides a more accurate estimation than the standard on/off test. Then, this result is transformed to the estimation of the initial characteristic equation parameters of the complex infinite-dimensional heat-exchanger model using the exact dominant-pole-loci assignment. The benefit of this technique is that multiple model parameters can be estimated under a single relay test. The second step attempts to estimate the remaining model parameters by various numerical optimization techniques and also to enhance all model parameters via the Autotune Variation Plus relay experiment for comparison. Although the obtained unordinary time and frequency domain responses may yield satisfactory results for control tasks, the identified model parameters may not reflect the actual values of process physical quantities.

Optimized implementation of an improved KNN classification algorithm using Intel FPGA platform: Covid-19 case study

The improved k-nearest neighbor (KNN) algorithm based on class contribution and feature weighting (DCT-KNN) is a highly accurate approach. However, it requires complex computational steps which consumes much time for the classification process. A field programmable gate array (FPGA) can be used to solve this drawback. However, using traditional hardware description language (HDL) to implement FPGA-based accelerators requires a high design time. Fortunately, the open computing language (OpenCL) high level parallel programming tool allows rapid and effective design on FPGA-based hardware accelerators. In this study, OpenCL has been used to examine speeding up the DCT-KNN algorithm on the FPGA parallel computing platform through applying numerous parallelization and optimization techniques. The optimized approach of the improved KNN could be used in various engineering problems that require a high speed of classification process. Classification of the COVID-19 disease is the case study used to examine this work. The experimental results show that implementing the DCT-KNN algorithm on the FPGA platform (Intel De5a-net Arria-10 device was used) gives an extremely high performance when compared to the traditional single-core-CPU based implementation. The execution time for our optimized design on the FPGA accelerator is 44 times faster than the conventional design implemented on the regular CPU-based computational platform.

Poplar optimization algorithm: A new meta-heuristic optimization technique for numerical optimization and image segmentation

A novel algorithm called Poplar Optimization Algorithm (POA) is developed in this paper to solve continuous optimization problems. The algorithm mimics the sexual and asexual propagation mechanism of poplar, where the basic philosophy of how to execute sexual and asexual propagation for individuals is detail designed in the algorithm. Mutation strategy of backtracking search algorithm is adopted in POA to maintain the diversity in a certain degree. The performance of POA algorithm is tested on 25 functions from the CEC2005 test suite and 30 functions from the CEC2017 test suite with different features. The results of POA are compared with some other population-based algorithms in terms of the quality and efficiency. Finally, the proposed algorithm is used to find the optimal threshold for image segmentation. The results indicate that the poplar optimization algorithm can obtain competitive or superior performance.

Optimal Resource Allocation in Economic Systems: A Numerical Approach

Optimal resource allocation is a critical challenge faced by modern manufacturing companies seeking to maximize profitability while adhering to resource limitations and market demands. This study investigates the application of numerical optimization techniques in addressing resource allocation complexities. Using a hypothetical case study of a manufacturing company producing electronic gadgets, we analyze the effectiveness of numerical methods in comparison to traditional analytical approaches. Through a detailed exploration of gradient-based optimization and derivative-free methods, we demonstrate the advantages of numerical optimization in achieving optimal resource allocations. Our findings underscore the significance of considering both demand forecasts and production capacities, leading to insightful policy recommendations for manufacturing companies and economic decision-makers. This research contributes to the field of economic optimization by highlighting the potential benefits of numerical approaches in enhancing resource allocation decision-making processes.

Mass production-enabled computational spectrometers based on multilayer thin films

Multilayer thin film (MTF) filter arrays for computational spectroscopy are fabricated using stencil lithography. The MTF filter array is a 6 × 6 square grid, and 169 identical arrays are fabricated on a single wafer. A computational spectrometer is formed by attaching the MTF filter array on a complementary metal–oxide–semiconductor (CMOS) image sensor. With a single exposure, 36 unique intensities of incident light are collected. The spectrum of the incident light is recovered using collected intensities and numerical optimization techniques. Varied light sources in the wavelength range of 500 to 849 nm are recovered with a spacing of 1 nm. The reconstructed spectra are a good match with the reference spectra, measured by a grating-based spectrometer. We also demonstrate computational pinhole spectral imaging using the MTF filter array. Adapting a spectral scanning method, we collect 36 monochromatic filtered images and reconstructed 350 monochromatic images in the wavelength range of 500 to 849 nm, with a spacing of 1 nm. These computational spectrometers could be useful for various applications that require compact size, high resolution, and wide working range.

Optimizing Friction Stir Welding of Dissimilar Grades of Aluminum Alloy Using WASPAS.

Aluminum is a widely popular material due to its low cost, low weight, good formability and capability to be machined easily. When a non-metal such as ceramic is added to aluminum alloy, it forms a composite. Metal Matrix Composites (MMCs) are emerging as alternatives to conventional metals due to their ability to withstand heavy load, excellent resistance to corrosion and wear, and comparatively high hardness and toughness. Aluminum Matrix Composites (AMCs), the most popular category in MMCs, have innumerable applications in various fields such as scientific research, structural, automobile, marine, aerospace, domestic and construction. Their attractive properties such as high strength-to-weight ratio, high hardness, high impact strength and superior tribological behavior enable them to be used in automobile components, aviation structures and parts of ships. Thus, in this research work an attempt has been made to fabricate Aluminum Alloys and Aluminum Matrix Composites (AMCs) using the popular synthesis technique called stir casting and join them by friction stir welding (FSW). Dissimilar grades of aluminum alloy, i.e., Al 6061 and Al 1100, are used for the experimental work. Alumina and Silicon Carbide are used as reinforcement with the aluminum matrix. Mechanical and corrosion properties are experimentally evaluated. The FSW process is analyzed by experimentally comparing the welded alloys and welded composites. Finally, the best suitable FSW combination is selected with the help of a Multi-Attribute Decision Making (MADM)-based numerical optimization technique called Weighted Aggregated Sum Product Assessment (WASPAS).

A New Approach to Optimize pH Buffers**

AbstractBuffer solutions are pervasive in chemistry, biochemistry, analytical chemistry, etc. A better understanding of buffer properties and what controls them is susceptible to be of interest in many scientific and technological fields. For instance, linear pH gradients are commonly used in electrophoresis and their optimization rests on numerical optimization of the concentrations of various weak species. It is probably generally assumed that no basic progress could be made on optimization approaches. We introduce here a new strategy to buffer optimization, based on a parametric study of the roots of the first derivative of the buffer index. In this way, it is possible to find mathematically optimal sets of parameters (pKa and concentrations). The method is applied to mixtures of 2, 3 and 4 monovalent species, which represent simple cases that do not call for overly elaborate numerical optimization techniques, but are nevertheless of practical interest in various branches of analytical chemistry.

Polarization Sensitive Imaging with Qubits

We compare reconstructed quantum state images of a birefringent sample using direct quantum state tomography and inverse numerical optimization technique. Qubits are used to characterize birefringence in a flat transparent plastic sample by means of polarization sensitive measurement using density matrices of two-level quantum entangled photons. Pairs of entangled photons are generated in a type-II nonlinear crystal. About half of the generated photons interact with a birefringent sample, and coincidence counts are recorded. Coincidence rates of entangled photons are measured for a set of sixteen polarization states. Tomographic and inverse numerical techniques are used to reconstruct the density matrix, the degree of entanglement, and concurrence for each pixel of the investigated sample. An inverse numerical optimization technique is used to obtain a density matrix from measured coincidence counts with the maximum probability. Presented results highlight the experimental noise reduction, greater density matrix estimation, and overall image enhancement. The outcome of the entanglement distillation through projective measurements is a superposition of Bell states with different amplitudes. These changes are used to characterize the birefringence of a 3M tape. Well-defined concurrence and entanglement images of the birefringence are presented. Our results show that inverse numerical techniques improve overall image quality and detail resolution. The technique described in this work has many potential applications.

Energy Efficiency of Inference Algorithms for Clinical Laboratory Data Sets: Green Artificial Intelligence Study.

BackgroundThe use of artificial intelligence (AI) in the medical domain has attracted considerable research interest. Inference applications in the medical domain require energy-efficient AI models. In contrast to other types of data in visual AI, data from medical laboratories usually comprise features with strong signals. Numerous energy optimization techniques have been developed to relieve the burden on the hardware required to deploy a complex learning model. However, the energy efficiency levels of different AI models used for medical applications have not been studied.ObjectiveThe aim of this study was to explore and compare the energy efficiency levels of commonly used machine learning algorithms—logistic regression (LR), k-nearest neighbor, support vector machine, random forest (RF), and extreme gradient boosting (XGB) algorithms, as well as four different variants of neural network (NN) algorithms—when applied to clinical laboratory datasets.MethodsWe applied the aforementioned algorithms to two distinct clinical laboratory data sets: a mass spectrometry data set regarding Staphylococcus aureus for predicting methicillin resistance (3338 cases; 268 features) and a urinalysis data set for predicting Trichomonas vaginalis infection (839,164 cases; 9 features). We compared the performance of the nine inference algorithms in terms of accuracy, area under the receiver operating characteristic curve (AUROC), time consumption, and power consumption. The time and power consumption levels were determined using performance counter data from Intel Power Gadget 3.5.ResultsThe experimental results indicated that the RF and XGB algorithms achieved the two highest AUROC values for both data sets (84.7% and 83.9%, respectively, for the mass spectrometry data set; 91.1% and 91.4%, respectively, for the urinalysis data set). The XGB and LR algorithms exhibited the shortest inference time for both data sets (0.47 milliseconds for both in the mass spectrometry data set; 0.39 and 0.47 milliseconds, respectively, for the urinalysis data set). Compared with the RF algorithm, the XGB and LR algorithms exhibited a 45% and 53%-60% reduction in inference time for the mass spectrometry and urinalysis data sets, respectively. In terms of energy efficiency, the XGB algorithm exhibited the lowest power consumption for the mass spectrometry data set (9.42 Watts) and the LR algorithm exhibited the lowest power consumption for the urinalysis data set (9.98 Watts). Compared with a five-hidden-layer NN, the XGB and LR algorithms achieved 16%-24% and 9%-13% lower power consumption levels for the mass spectrometry and urinalysis data sets, respectively. In all experiments, the XGB algorithm exhibited the best performance in terms of accuracy, run time, and energy efficiency.ConclusionsThe XGB algorithm achieved balanced performance levels in terms of AUROC, run time, and energy efficiency for the two clinical laboratory data sets. Considering the energy constraints in real-world scenarios, the XGB algorithm is ideal for medical AI applications.

Design specification management with automated decision-making for reliable optimization of miniaturized microwave components

The employment of numerical optimization techniques for parameter tuning of microwave components has nowadays become a commonplace. In pursuit of reliability, it is most often carried out at the level of full-wave electromagnetic (EM) simulation models, incurring considerable computational expenses. In the case of miniaturized microstrip circuits, densely arranged layouts with strong cross-coupling effects make EM-driven tuning imperative to achieve the optimum performance. The process is even more challenging due to a typically large number of geometry parameters, and the lack of reasonable initial designs. The latter often encourages the use of global search procedures, which may be prohibitively expensive. In this paper, a novel automated framework for reliable optimization of miniaturized microwave components is proposed. Our methodology is based on design specification management, where the performance requirements imposed on the system are temporarily relaxed if the current design is unlikely to be improved (e.g., due to being away from the target operating frequency). The specifications are re-adjusted at each iteration of the algorithm, and eventually converge to their original values. Using two examples of compact microstrip couplers and a power divider, the presented technique is demonstrated to significantly improve the efficacy of local search routines under challenging design scenarios.

Optimization of CA-TIG welding parameters to predict and maximize tensile properties of super alloy 718 sheets for gas turbine applications

PurposeThe primary objective of this investigation is to optimize the constricted arc tungsten inert gas (CA-TIG) welding parameters specifically welding current (WC), arc constriction current (ACC), ACC frequency (ACCF) and CA traverse speed to maximize the tensile properties of thin Inconel 718 sheets (2 mm thick) using a statistical technique of response surface methodology and desirability function for gas turbine engine applications.Design/methodology/approachThe four factor – five level central composite design (4 × 5 – CCD) matrix pertaining to the minimum number of experiments was chosen in this investigation for designing the experimental matrix. The techniques of numerical and graphical optimization were used to find the optimal conditions of CA-TIG welding parameters.FindingsThe thin sheets of Inconel 718 (2 mm thick) can be welded successfully using CA-TIG welding process without any defects. The joints welded using optimized conditions of CA-TIG welding parameters showed maximum of 99.20%, 94.45% and 73.5% of base metal tensile strength, yield strength and elongation.Originality/valueThe joints made using optimized CA-TIG welding parameters disclosed 99.20% joint efficiency which is comparatively 20%–30% superior than conventional TIG welding process and comparable to costly electron beam welding and laser beam welding processes. The parametric mathematical equations were designed to predict the tensile properties of Inconel 718 joints accurately with a confidence level of 95% and less than 4.5% error. The mathematical relationships were also developed to predict the tensile properties of joints from the grain size (secondary dendritic arm spacing-SDAS) of fusion zone microstructure.

Group‐level information decay and inventory inspection: An empirical‐analytical approach

AbstractDespite numerous optimization techniques proposed to address inventory record inaccuracy (IRI), there continues to be a lack of empirical appreciation of IRI, nor do existing item‐level inspection policies accommodate managerial needs for allocating inspection effort to groups of items. We address these issues by collaborating with a retailer to collect archival data and develop statistical estimation and practical optimization techniques that address IRI. We propose a continuous‐time model to explain group‐level information decay and, based on the model's structure, develop an inspection policy that satisfies managerial needs for allocating inspection effort to a group of items. After analyzing model behavior and quantifying unobservable cost factors using information readily available, we test the efficacy of the proposed policy within the retailing environment and find it to outperform both the retailer's inspection routines and a well‐established industry benchmark. Finally, we illustrate how to optimize grouping and inspection decisions jointly and show potential benefits from optimized group formation. Our empirically grounded modeling effort yields transferable methods and generalizable insights for improving inventory integrity.

Comparison among limit performances of various chassis control methods by numerical optimization

Since the limit performance of an integrated chassis control vehicle depends on the combination of control inputs such as steering angle, braking torque, and so on, it is important to clarify how the combination of control inputs affects the limit performance. In this study, the performances of integrated chassis control systems with various input combinations are evaluated in an obstacle avoidance situation. The obstacle avoidance problem is formulated as an optimal control problem that maximizes the avoidable obstacle width at the predefined distance and is solved by using a numerical optimization technique. In the optimization, a full-vehicle model equipped with actuators that can independently control the steering angles, braking torques, and active suspension forces of four wheels is used, and the optimization results are compared for various combinations of control inputs. From the optimization results for four combinations of control inputs, the effects of integrated control are examined.

GELAB – The Cutting Edge of Grammatical Evolution

The advent of cloud-based super-computing platforms has given rise to a Data Science (DS) boom. Many types of technological problems that were once considered prohibitively expensive to tackle are now candidates for exploration. Machine Learning (ML) tools that were valued only in academic environments are quickly being embraced by industrial giants and tiny startups alike. Coupled with modern-day computing power, ML tools can be looked at as hammers that can deal with even the most stubborn nails. ML tools have become so ubiquitous that the current industrial expectation is that they should not only deliver accurate and intelligent solutions but also do so rapidly. In order to keep pace with these requirements, a new enterprise, referred to as MLOps has blossomed in recent years. MLOps combines the process of ML and DS with an agile software engineering technique to develop and deliver solutions in a fast and iterative way. One of the key challenges to this is that ML and DS tools should be efficient and have better usability characteristics than were traditionally offered. In this paper, we present a novel software for Grammatical Evolution (GE) that meets both of these expectations. Our tool, GELAB, is a toolbox for GE in Matlab which has numerous features that distinguish it from existing contemporary GE software. Firstly, it is user-friendly and its development was aimed for use by non-specialists. Secondly, it is capable of hybrid optimization, in which standard numerical optimization techniques can be added to GE. We have shown experimentally that when hybridized with meta-heuristics GELAB has an overall better performance as compared with standard GE.

Numerical optimisation technique to solve imprecisely defined nonlinear system of equations with bounded parameters

Numerical optimisation technique to solve imprecisely defined nonlinear system of equations with bounded parameters

Auditing and Debugging Deep Learning Models via Flip Points: Individual-Level and Group-Level Analysis

Deep learning models have been criticized for their lack of easy interpretation, which undermines confidence in their use for important applications. Nevertheless, they are consistently utilized in many applications, consequential to humans’ lives, usually because of their better performance. Therefore, there is a great need for computational methods that can explain, audit, and debug such models. Here, we use flip points to accomplish these goals for deep learning classifiers used in social applications. A trained deep learning classifier is a mathematical function that maps inputs to classes. By way of training, the function partitions its domain and assigns a class to each of the partitions. Partitions are defined by the decision boundaries which are expected to be geometrically complex. This complexity is usually what makes deep learning models powerful classifiers. Flip points are points on those boundaries and, therefore, the key to understanding and changing the functional behavior of models. We use advanced numerical optimization techniques and state-of-the-art methods in numerical linear algebra, such as rank determination and reduced-order models to compute and analyze them. The resulting insight into the decision boundaries of a deep model can clearly explain the model’s output on the individual level, via an explanation report that is understandable by non-experts. We also develop a procedure to understand and audit model behavior towards groups of people. We show that examining decision boundaries of models in certain subspaces can reveal hidden biases that are not easily detectable. Flip points can also be used as synthetic data to alter the decision boundaries of a model and improve their functional behaviors. We demonstrate our methods by investigating several models trained on standard datasets used in social applications of machine learning. We also identify the features that are most responsible for particular classifications and misclassifications. Finally, we discuss the implications of our auditing procedure in the public policy domain.

MIMO Hybrid FSO/RF System Over Generalized Fading Channels

Free-space optics (FSO) communication is often unreliable, especially for long distances, as it suffers from attenuation due to foggy weather, atmospheric turbulence, and pointing errors. To avoid communication link failure, radio frequency (RF) communication, which is unaffected from these effects, is utilized when FSO communication is in outage. Diversity is known to counteract the adverse fading effects, therefore, this paper presents a transmit aperture/antenna selection (TAS) and selection combining (SC)-based multiple-input-multiple-output (MIMO) hybrid FSO/RF system. The proposed system consists of a MIMO FSO and RF sub-systems. The MIMO FSO sub-system is given higher priority to transmit and receive using the selected best link. The best link of the MIMO RF sub-system is used when the FSO sub-system is in outage. The probability density function (PDF) and cumulative distribution function (CDF) expressions for TAS/SC-based MIMO FSO and RF sub-systems are derived. The derived PDF and CDF expressions are used further to derive the exact and asymptotic expressions for outage probability, average symbol error rate (SER), and capacity of the hybrid system. From the asymptotic SER expressions, the diversity order of the proposed system is determined. The optimum values for the threshold SNR and transmit beam waist, which minimize the average SER, are obtained. Further, the obtained optimal values using analytical expressions are verified using numerical optimization technique. The effect of multiple apertures/antennas, pointing errors, and atmospheric attenuation over the system performance is analyzed extensively by varying the FSO and RF channel parameters. The results show that the proposed hybrid FSO/RF system is more reliable compared to the individual MIMO FSO and RF systems and achieves higher capacity compared to the MIMO RF system.