MATLAB Code Research Articles

Modelowe i empiryczne podejście do oceny parametrów petrofizycznych skał na przykładzie złoża gazu Dhodak w Pakistanie

The main aim of the presented study was to perform an analysis of petrophysical rock parameters for the Pab Formation, which is a potential reservoir, located at four-depth intervals within the Dhodak Gas Field. The Pab Formation in the studied wells is mostly composed of the well-sorted sandstone. The intervals for the study were selected by applying a gamma ray log cut-off of 40%. A Matlab code that is user-friendly and contains empirical and numerical models (Raymor’s, Han's, Castagna’s, Jizba, Blangy, and Gardner's models) was designed with the purpose of providing an interpretation of rock physics parameters in reservoirs. The results indicate that the Pab Formation is mainly saturated with hydrocarbons and possesses sufficient porosity (i.e. between 8 and 38%). According to the cross-plot models, there are a few zones within the Pab Formation that are weakly consolidated. These zones are located at depths of 2377–2404 m and 2418–2430 m. However, the same area is classified as tight gas sandstone and is located between 2330 to 2343 m. The results of the study indicate that the Pab Formation is potentially saturated with hydrocarbons. These findings were based on structural and rock physics assessments of the study area. The approach presented in this work is also helpful for structural trap evaluation. In addition, the petrophysical rock parameter models developed can contribute to reducing the risks associated with hydrocarbon exploration, enabling accurate representation of subsurface conditions and improving reservoir forecasting capabilities in oil producing areas.

Condition monitoring of a CNC hobbing cutter using machine learning approach

The state of cutting tools profoundly influences the efficiency of the machining processes within the manufacturing industry. Cutting tool faults are highly undesirable and can adversely impact the performance of machine tools, leading to a shortened operational lifespan. Consequently, it is imperative to minimize power consumption by closely monitoring the condition of cutting tools. This necessitates implementing an effective supervision system to continually assess and predict potential faults. In simpler terms, this entails identifying issues that could compromise the lifespan of cutting tools before they escalate into problems like wear, breakage, or complete failure. This proactive approach ensures the optimal and efficient utilization of cutting tools, reduces the need for maintenance and repair, and enhances process stability, among other benefits. In this paper, a novel approach of machine learning for the condition monitoring of hobbing cutters used in Computer Numerical Control Machines (CNCs) was built. Vibration signals from hobbing blades were recorded under various conditions, including healthy and faulty states. The histogram presents a comprehensive overview of the statistical distribution for five variables related to the studied dataset. MATLAB code and scripts are utilized for extracting relevant statistical features, and for identification of the most relevant features, decision tree algorithms were used. For training the ML algorithms the hyperparameters were selected by the Grid Search Method and the Principal component analysis (PCA) was enabled for the reduction of dimensionality and to simplify the data set. The various conditions of hobbing cutters were then classified using tree-based classification models, giving 100% classification accuracy. It helps to develop a novel condition monitoring system for CNC hobbing cutters using machine learning methods to identify problems in hobbing blades. This would ultimately lead to lower power consumption and enhanced performance of machine tools.

Fat-free noncontrast whole-heart CMR with fast and power-optimized off-resonant water excitation pulses

BackgroundCardiovascular MRI (CMR) faces challenges due to the interference of bright fat signals in visualizing structures like coronary arteries. Effective fat suppression is crucial, especially when using whole-heart CMR techniques. Conventional methods often fall short due to rapid fat signal recovery, leading to residual fat content hindering visualization. Water-selective off-resonant radiofrequency (RF) pulses have been proposed but come with tradeoffs between pulse duration, which increases scan time, and increased RF energy deposit, which limits their applicability due to specific absorption rate (SAR) constraints. The study introduces a lipid-insensitive binomial off-resonant (LIBOR) RF pulse, which addresses concerns about SAR and scan time, and aims to provide a comprehensive quantitative comparison with published off-resonant RF pulses for CMR at 3T. MethodsA short (1ms) LIBOR pulse, with reduced RF power requirements, was developed and implemented in a free-breathing respiratory-self-navigated 3D radial whole-heart CMR sequence at 3T. A binomial off-resonant rectangular (BORR) pulse with matched duration, as well as previously published lipid-insensitive binomial off-resonant excitation (LIBRE) pulses (1ms and 2.2ms), were implemented and optimized for fat suppression in numerical simulations and validated in volunteers (n=3). Whole-heart CMR was performed in volunteers(n=10) with all four pulses. The signal-to-noise ratio (SNR) of ventricular blood, skeletal muscle, myocardium, and subcutaneous fat and the coronary vessel detection rates and sharpness were compared. ResultsExperimental results validated numerical findings and near homogeneous fat suppression was achieved with all four pulses. Comparing the short RF pulses (1ms), LIBOR reduced the RF power nearly two-fold compared with LIBRE, and three-fold compared with BORR, and LIBOR significantly decreased overall fat SNR from cardiac scans, compared to LIBRE and BORR. The reduction in RF pulse duration (from 2.2ms to 1ms) shortened the whole-heart acquisition from 8.5min to 7min. No significant differences in coronary arteries detection and sharpness were found when comparing all four pulses. ConclusionLIBOR pulses enabled whole-heart CMR under 7minutes at 3T, with large volume fat signal suppression, while reducing RF power compared with LIBRE and BORR pulses. LIBOR is an excellent candidate to address SAR problems encountered in CMR sequences where fat suppression remains challenging and short RF pulses are required. AVAILABILITY OF DATA AND MATERIALSAn online repository containing the anonymized human MRI raw data, as well as RF pulse shapes used in this study is publicly available at:https://zenodo.org/records/8338079(PART 1: KNEE V1-V3, HEART V1-V5)https://zenodo.org/records/10715769 (PART 2: HEART V6-V10)Matlab code to 1) simulate the different RF pulses within a GRE sequence and 2) to read and display the anonymized raw data is available from: https://github.com/QIS-MRI/LIBOR_LIBRE_BORR_SimulationCodeThe compiled research sequence can be requested through the Teamplay platform of Siemens Healthineers.

Multi-epitope vaccine design of African swine fever virus considering T cell and B cell immunogenicity

T and B cell activation are equally important in triggering and orchestrating adaptive host responses to design multi-epitope African swine fever virus (ASFV) vaccines. However, few design methods have considered the trade-off between T and B cell immunogenicity when identifying promising ASFV epitopes. This work proposed a novel Pareto front-based ASFV screening method PFAS to identify promising epitopes for designing multi-epitope vaccines utilizing five ASFV Georgia 2007/1 sequences. To accurately predict T cell immunogenicity, four scoring methods were used to estimate the T cell activation in the four stages, including proteasomal cleavage probability, transporter associated with antigen processing transport efficiency, class I binding affinity of the major histocompatibility complex, and CD8 + cytotoxic T cell immunogenicity. PFAS ranked promising epitopes using a Pareto front method considering T and B cell immunogenicity. The coefficient of determination between the Pareto ranks of multi-epitope vaccines and survival days of swine vaccinations was R2 = 0.95. Consequently, PFAS scored complete epitope profiles and identified 72 promising top-ranked epitopes, including 46 CD2v epitopes, two p30 epitopes, 10 p72 epitopes, and 14 pp220 epitopes. PFAS is the first method of using the Pareto front approach to identify promising epitopes that considers the objectives of maximizing both T and B cell immunogenicity. The top-ranked promising epitopes can be cost-effectively validated in vitro. The Pareto front approach can be adaptively applied to various epitope predictors for bacterial, viral and cancer vaccine developments. The MATLAB code of the Pareto front method was available at https://github.com/NYCU-ICLAB/PFAS.Graphical

Does selective attention demand affect the physiological complexity of postural control in faller community-dwelling older adults?

Background: The physiological complexity of postural control during selective attention demand can be used as an innovative method to identify faller community-dwelling older. Objective: To analyze the effect of selective attention on the complexity of postural control in faller community-dwelling older adults. Methods: A total of 57 older adults (60 to 80 years) were divided between the faller group (Fallers n = 21) and the non-faller group (Non-fallers n = 36). An inertial sensor (Physilog® 5, GaitUp, Switzerland) was positioned over the second sacral vertebra to collect the participant’s body sway in the anteroposterior (AP) and mediolateral (ML) directions during the monochromatic and color Stroop test versions. The postural control’s complexity was calculated using MATLAB codes, employing the refined composite multiscale fuzzy entropy method. Statistical Package for Social Sciences was used to analyze the effects of interaction, group, and condition effects using linear mixed models with an alpha of 5%. Results: No significant interaction effects were observed in the AP (F = 0.18; p = 0.66) and ML (F = 0.00; p = 0.99) directions. No group effect was observed in the AP (F = 1.23; p = 0.26) and ML (F = 1.76; p = 0.18) directions. No condition effect was found in the AP (F = 0.06; p = 0.80) and ML (F = 3.54; p = 0.06) directions. Conclusion: Once-only faller community-dwelling older adults did not evidence worse physiological complexity than the non-fallers during selective attention demand in the upright standing posture.

Designing a TPMS metamaterial via deep learning and topology optimization

Data-driven models that act as surrogates for computationally costly 3D topology optimization techniques are very popular because they help alleviate multiple time-consuming 3D finite element analyses during optimization. In this study, one such 3D CNN-based surrogate model for the topology optimization of Schoen’s gyroid triply periodic minimal surface unit cell is investigated. Gyroid-like unit cells are designed using a voxel algorithm and homogenization-based topology optimization codes in MATLAB. A few such optimization data are used as input–output for supervised learning of the topology-optimization process via the 3D CNN model in Python code. These models could then be used to instantaneously predict the optimized unit cell geometry for any topology parameters. The high accuracy of the model was demonstrated by a low mean square error metric and a high Dice coefficient metric. The model has the major disadvantage of running numerous costly topology optimization runs but has the advantages that the trained model can be reused for different cases of TO and that the methodology of the accelerated design of 3D metamaterials can be extended for designing any complex, computationally costly problems of metamaterials with multi-objective properties or multiscale applications. The main purpose of this paper is to provide the complete associated MATLAB and PYTHON codes for optimizing the topology of any cellular structure and predicting new topologies using deep learning for educational purposes.

A general method for mode decomposition on additive mixture: Generalized Variational Mode Decomposition and its sequentialization

Variational Mode Decomposition(VMD) method was proposed to separate non-stationary signal mixture by solving a optimization problem. This method is powerful and can reconstruct the signal components precisely when they are orthogonal(or quasi-orthogonal) in frequency domain. The crucial problem for VMD is that it requires the information of modal number before the decomposition. Also its applications are limited in 1D and 2D signal processing fields, of narrow scope.In this paper, by inheriting and developing the core idea of VMD, we build a general form for this method and extend it to the modal decomposition for common additive mixture, not only limited in signal processing. To overcome the obstacle of modal number, we sequentialize the generalized VMD method, such that the modes can be extracted one by one, without knowing the modal number a priori. After the generalization and sequentialization for the VMD, we apply them in different fields of additive case, such as texture segmentation, Gaussian Mixture Model(GMM), clustering, etc. From the experiments, we conclude that the generalized and sequentialized VMD methods can solve variety classical problems from the view of modal decomposition, which implies that our methods have higher generality and wider applicability. A raw Matlab code for this algorithm is shown in https://github.com/changwangke/SGVMD_additive_Clustering/blob/main/SGVMD_clustering.m.

A simple graphical method for sizing flat‐plate solar air heaters based on transversal and longitudinal aspect ratios

AbstractSolar air heaters (SAHs) are widely used for drying vegetables and fruits or for domestic heating. Certain sizing parameters are necessary to obtain the right dimensions for the required air temperature, flow rate, and thus useful thermal energy. The well‐known Hottel–Whillier–Bliss equation was used and made dimensionless by applying the collector longitudinal and transversal aspect ratios (rl and rt) of a double‐glazed flat plate solar air heater (DG‐FPSAH). The steady‐state equations are solved to determine the average temperatures. Thereafter, one could calculate the overall loss heat coefficient and efficiency factor, obtained analytically. A Matlab code was developed to estimate primarily unknown temperatures, useful energy, and the Nusselt number. An iterative numerical method is used until convergence occurs. The inlet cross‐sectional area and air flow velocity are defined as input data. The proposed sizing method depends on the output temperature required by the customer. This temperature can be determined from the plotted curves of the dimensionless ratios. Hence, the SAH‐needed dimensions are determined graphically depending on the functional requirements for construction planning, such as technology choice, work breakdown, and budgeting. In the present case study, based on the input parameters, an airflow rate of 1.2 kg/s entering a DG‐FPSAH with an output temperature of 41.5°C yields the dimensions of Lin = 3.824 m, Win = 2.735 m, and Hin = 0.1825, specifying the collector length, width, and air duct height. The gathered energy and thermohydraulic efficiency are: Qu = 7.91 kW, ηcol = 0.752, respectively.

Impact of north-west Bay of Bengal bathymetry along the off-coast of Digha on the propagation of shallow water waves using finite difference method

Numerical simulation of shallow water equations has been carried out using the Crank-Nicolson scheme of Finite Difference Method (FDM) to study the interaction of waves with different forms of bathymetry especially that of the off-coast of Digha, a tourist spot situated in the north-west coast of Bay of Bengal. MATLAB code has been developed for the simulation. One-dimensional (1-D) linear shallow water model has been considered here. Three different bathymetry profile are chosen for the investigation: (i) flat bathymetry (with or without seamount/basin); (ii) piecewise linear approximation to the real bathymetry of the off-coast of Digha; (iii) real bathymetry of the off-coast of Digha. Real bathymetry data has been obtained from the gridded data set of GEBCO (General Bathymetric Chart of the Oceans) published in 2022. Validation of the present numerical scheme has been performed by considering test problems with results existing in the literature. Comparative study of the effect of the bathymetries, on the propagation of shallow water waves, has been performed and demonstrated with figures and tables. Results of wave-bathymetry interaction obtained here are believed to be new with respect to the present study area and will definitely give some insight into the phenomena in this region.

Algorithm Validation for Quantifying ActiGraph™ Physical Activity Metrics in Individuals with Chronic Low Back Pain and Healthy Controls.

Assessing physical activity is important in the treatment of chronic conditions, including chronic low back pain (cLBP). ActiGraph™, a widely used physical activity monitor, collects raw acceleration data, and processes these data through proprietary algorithms to produce physical activity measures. The purpose of this study was to replicate ActiGraph™ algorithms in MATLAB and test the validity of this method with both healthy controls and participants with cLBP. MATLAB code was developed to replicate ActiGraph™'s activity counts and step counts algorithms, to sum the activity counts into counts per minute (CPM), and categorize each minute into activity intensity cut points. A free-living validation was performed where 24 individuals, 12 cLBP and 12 healthy, wore an ActiGraph™ GT9X on their non-dominant hip for up to seven days. The raw acceleration data were processed in both ActiLife™ (v6), ActiGraph™'s data analysis software platform, and through MATLAB (2022a). Percent errors between methods for all 24 participants, as well as separated by cLBP and healthy, were all less than 2%. ActiGraph™ algorithms were replicated and validated for both populations, based on minimal error differences between ActiLife™ and MATLAB, allowing researchers to analyze data from any accelerometer in a manner comparable to ActiLife™.

Vector to matrix representation for CNN networks for classifying astronomical data

Choosing the right classifier is crucial for effective classification in various astronomical datasets aimed at pattern recognition. While the literature offers numerous solutions, the support vector machine (SVM) continues to be a preferred choice across many scientific fields due to its user-friendliness. In this study, we introduce a novel approach using convolutional neural networks (CNNs) as an alternative to SVMs. CNNs excel at handling image data, which is arranged in a grid pattern. Our research explores converting one-dimensional vector data into two-dimensional matrices so that CNNs pre-trained on large image datasets can be applied. We evaluate different methods to input data into standard CNNs by using two-dimensional feature vector formats. In this work, we propose a new method of data restructuring based on a set of wavelet transforms. The robustness of our approach is tested across two benchmark datasets/problems: brown dwarf identification and threshold crossing event (Kepler data) classification. The proposed ensembles produce promising results on both datasets. The MATLAB code of the proposed ensemble is available at https://github.com/LorisNanni/Vector-to-matrix-representation-for-CNN-networks-for-classifying-astronomical-data

RBE-based dose planning, and calculation of TCP and NTCP with gold nanoparticles for intermediate photon energy in pancreatic cancer

Objective. This study simulated the potential of gold nanoparticles (GNPs) to improve the effectiveness of radiation therapy in pancreatic cancer cases. The purpose of this study was to assess the impact of GNPs on tumor control probability (TCP) and normal tissue complication probability (NTCP) in pancreatic cancer cases undergoing radiation therapy. The work aimed to compare treatment plans generated with a novel 2.5 MV beam using GNPs to conventional 6 MV plans and evaluate the dose-volume histogram (DVH), TCP, and NTCP. Approach. Treatment planning for five pancreatic computed tomography (CT) images was performed using the open-source MATLAB-based treatment planning program matRad. MATLAB codes were developed to calculate the relative biological effectiveness (RBE) of GNPs and apply the corresponding dose and RBE values to each voxel. TCP and NTCP were calculated based on the applied RBE values. Main results. Adding GNPs to the 2.5 MV treatment plan resulted in a significant increase in TCP, from around 59% to 93.5%, indicating that the inclusion of GNPs improved the effectiveness of the radiation treatment. The range in NTCP without GNPs was relatively larger compared to that with GNPs. Significance. The results indicated that the addition of GNPs to a 2.5 MV plan can increase TCP while maintaining a relatively low NTCP value (<1%). The use of GNPs may also reduce NTCP values by decreasing the dose to normal tissues while maintaining the same prescribed dose to the tumor. Hence, the addition of GNPs can improve the balance between TCP and NTCP.

Numerical investigation of the effects of adding different gases on the performance and emissions of an ammonia-hydrogen dual-fuel engine

As an alternative to fossil fuels, there is growing interest in using ammonia in combustion systems, particularly in internal combustion engines. As a competent competitor to hydrogen, it has various advantages. However, adding an ignition promoter such as hydrogen is still necessary to maintain combustion stability. Since the engine using ammonia also suffers from high NOx emissions, in this study, the effects of adding different gases on the performance and emissions of an ammonia-hydrogen dual-fuel engine were numerically investigated. The base engine was a diesel engine whose parameters were accustomed to running with ammonia-hydrogen as fuel. Hydrogen was injected via the port in the intake stage at 180 crank angle degrees (CAD) and mixed with the cylinder charge. Ammonia was directly injected into the cylinder at 350–370 CAD. Different gasses, including argon, nitrogen, carbon dioxide, and oxygen, were injected into the cylinder at various crank angles before, during, and after ammonia injection (330–350 CAD, 350–370 CAD, and 370–390 CAD). A MATLAB code was prepared to solve the governing equations, and the combustion mechanism was implemented in Cantera. The results showed that adding CO2 before or concurrent with the ammonia injection timing had undesirable impacts on the peak in-cylinder pressure and NO and NO2 emissions. The O2 addition had a negative on the emissions. Adding N2 and Ar concurrently with the ammonia injection (350–370 CAD) could diminish NO and NO2 emissions without drastically affecting the peak in-cylinder pressure and temperature.

The effects of bioconvection, non‐Fourier heat flux, and thermal radiations on Williamson nanofluids and Maxwell nanofluids transportation with prescribed thermal conditions

AbstractThe utilization of nanoentities in common fluids has opened new opportunities in the area of heat transportation. The rising requirements to enhance the efficiency of compact heat exchangers can be achieved by using various nanofluids. In this article, the thermal output of Maxwell and Williamson nanofluids transport over a prolonging sheet with bioconvection of self‐motivated organisms is scrutinized. A magnetic flux and the porous effects of a medium influence the flow of fluids. The fundamental principles for conservation of mass, concentration, momentum, and energy yield a nonlinear set of partial differential equations that can then be altered into ordinary differential form. A heat transfer flux is presented along with temperature boundary conditions, PST, and PHF (prescribed surface temperature and prescribed heat flux). The numerical results are acquired by executing the Runge–Kutta method with a shooting procedure in MATLAB coding. By fluctuating the inputs of influential variables of the dependent functions, a precise overview of the scheme is acquired. It can be seen that velocity decreases with the rising values of buoyancy ratio, magnetic force, Raleigh number, and porosity. Also, the temperature of the fluids begins to rise directly with the rising values of thermophoresis and Brownian motion parameters. The present study addresses bioconvection, non‐Fourier heat flow, and thermal radiations while combining the special properties of Williamson and Maxwell nanofluids. The field of biomedical engineering may benefit from this study, particularly with regard to therapies for hyperthermia and drug delivery systems. This study can be useful in cutting‐edge cooling systems, bioengineering, solar energy conversion and biotechnology.

Exploring the influence of morphology on magnetized Ree–Eyring tri‐hybrid nanofluid flow between orthogonally moving coaxial disks using artificial neural networks with Levenberg–Marquardt scheme

AbstractThe present study presents an analysis of Ree–Eyring tri‐hybrid nanofluid flow between two expanding/contracting disks with permeable walls by applying the computing power of Levenberg–Marquardt supervised neural networks (LM‐SNNs). The effects of thermal radiation, Brownian motion, and thermophoresis were also thoroughly examined. The results are presented for tri‐hybrid nanofluid with SWCNT and MWCNT and Fe2O3 and H2O base fluid. The coupled non‐linear PDE system is transformed into a system of ODE associated with convective boundary conditions by applying the appropriate transformations. This is then accomplished numerically by using the finite difference‐based BVP‐4c MATLAB code that implements the three‐stage Lobatto IIIA formula. The results are novel and have been validated with LM‐SNNs outcomes. It has been observed that both numerical outcomes and LM‐SNNs produce equivalent results, and both approaches exhibit a drop in the velocity profile for the magnetic field near the lower plate and a rise near the upper plate. The skin friction against the Prandtl number increases, whereas the Nusselt number decreases at the upper disc. Compared to BVP‐4c numerical approaches, the given LM‐SNNs model is more dependable, efficient, and time‐saving because it requires less work and produces results quickly.

X-IGA Used for Orthotropic Material Crack Growth.

In this paper, we propose a new approach for numerically simulating the growth of cracks in unidirectional composite materials, termed extended isogeometric analysis, evaluating the maximum stress intensity factor and T-stress. To validate our approach, we used a small anisotropic plate with two edge cracks, beginning with formulating the governing equations based on the energy integral method, Stroh's Formula, and the Elastic Law describing the behaviour of anisotropic materials, while considering boundary conditions and initial states. A MATLAB code was developed to solve these equations numerically and to post-process the tensile stress and the stress intensity factor (SIF) in the first mode. The results for the SIF closely match those obtained using the extended finite element method (X-FEM), with a discrepancy of only 0.0021 Pa·m0.5. This finding underscores the credibility of our approach. The extended finite element method has demonstrated robustness in predicting crack propagation in composite materials in recent years, leading to its adoption by several widely used software packages in various industries.

Application of fractal analysis to characterise the fracture and mechanical properties of concrete.

Abstract Concrete structures consist of cracks even under service loads. The presence of micro-cracks and other flaws such as pores and voids act as potential sources for crack propagation leading to fracture under external loadings. Irregularities produced during the fracturing process in concrete are manifested on the fracture surface, characterizing it as fractal. Therefore, the investigation and quantification of the fractal pattern exhibited by fracture surfaces have become a focal point of research in recent years. In this work, beam specimens made with four different concrete mixes underwent a three-point bending test in order to determine a correlation between the fracture energy and mechanical characteristics of concrete. Concrete’s mechanical characteristics and fracture energy are discovered to be positively correlated. The fractal analysis is done using an LVDT instrument. Using a program written in MATLAB code, the fractured surface is regenerated with measured undulations. Concrete’s flexural strength and fracture energy are then connected with the fractal dimension value that was found. In conclusion, it was observed that a strong correlation exists between the fractal dimension value and fracture energy.

Determination of the effects of size parameters and lean angle on Bovet-based Francis turbine runner performance

Abstract In global hydropower generation, the Francis turbine is widely used because of its ability to work efficiently in a wide range of conditions like varying head and flow rates. Turbine’s meridional profile can be developed using traditional design approaches, like the Bovet method which uses empirical relations based on hydrological factors, that can be refined through optimization. Runner performance evaluation considers factors such as efficiency, erosion resistance, resistance to cavitation, and many more. Adjusting design parameters can enhance a turbine’s performance at a particular site. In this paper, a Bovet-based Francis runner design was obtained for Damile Micro Hydro located at Pharping, Nepal and the effect of variation of size parameters and lean angle on the runner performance was investigated. Initially, In-house MATLAB codes were used to design 99 meridional profile of different sizes by changing the size parameters as indicated by Bovet method and numerical simulation was performed to select the best meridional profile based on efficiency. After obtaining best meridional profile, the effect of lean angle was studied by varying it from -10° to 10°. Among the 99 designs based on different sizes, the simulation results indicate that the maximum efficiency was found to be 92.3% whereas the minimum efficiency was found to be 87.84%. The corrected lean angle of -5° showed the maximum efficiency and better pressure distribution along the blade.

Randomized algorithms for large-scale dictionary learning

Dictionary learning is an important sparse representation algorithm which has been widely used in machine learning and artificial intelligence. However, for massive data in the big data era, classical dictionary learning algorithms are computationally expensive and even can be infeasible. To overcome this difficulty, we propose new dictionary learning methods based on randomized algorithms. The contributions of this work are as follows. First, we find that dictionary matrix is often numerically low-rank. Based on this property, we apply randomized singular value decomposition (RSVD) to the dictionary matrix, and propose a randomized algorithm for linear dictionary learning. Compared with the classical K-SVD algorithm, an advantage is that one can update all the elements of the dictionary matrix simultaneously. Second, to the best of our knowledge, there are few theoretical results on why one can solve the involved matrix computation problems inexactly in dictionary learning. To fill-in this gap, we show the rationality of this randomized algorithm with inexact solving, from a matrix perturbation analysis point of view. Third, based on the numerically low-rank property and Nyström approximation of the kernel matrix, we propose a randomized kernel dictionary learning algorithm, and establish the distance between the exact solution and the computed solution, to show the effectiveness of the proposed randomized kernel dictionary learning algorithm. Fourth, we propose an efficient scheme for the testing stage in kernel dictionary learning. By using this strategy, there is no need to form nor store kernel matrices explicitly both in the training and the testing stages. Comprehensive numerical experiments are performed on some real-world data sets. Numerical results demonstrate the rationality of our strategies, and show that the proposed algorithms are much efficient than some state-of-the-art dictionary learning algorithms. The MATLAB codes of the proposed algorithms are publicly available from https://github.com/Jiali-yang/RALDL_RAKDL.

Crack growth in sandwich-structured foam core graphite epoxy laminate composite using a phase-field modelling approach

The laminated sandwich composites have wide structure-making applications in the automotive and aviation fields due to their lightweight and superior flexural rigidity properties. Making grooves or holes to assemble more than one structure induces crack discontinuities near the stress concentration region in these sandwich structures. The present work examines the effect of crack discontinuities on the mechanical performance and failure process of the sandwich structures under different loading conditions. Phase field method (PFM) has been presented and implemented using in-house developed MATLAB code. The effect of holes, multiple cracks, number of cores, and loading conditions are analyzed for the mechanical and fracture behavior of the structure. Load-carrying capacity, threshold displacement value for crack initiation, crack propagation trajectory, and energy absorption capacity are compared for various crack discontinuities under different loading conditions. Approximately 35% increase in load carrying capacity is observed in equivalent multiple core sandwich structures.