MATLAB Parallel Computing Toolbox Research Articles

Technical note: MAMBA-Multi-pAradigM voxel-Based Analysis: A computational cookbot.

Voxel-Based (VB) analysis embraces a multifaceted ensemble of sophisticated techniques, lying at the boundary between image processing and statistical modeling, that allow for a frequentist inference of pathophysiological properties anchored to an anatomical reference. VB methods has been widely adopted in neuroimaging studies and, more recently, they are gaining momentum in radiation oncology research. However, the price for the power of VB analysis is the complexity of the underlying mathematics and algorithms. In this paper, we present the Multi-pAradigM voxel-Based Analysis (MAMBA) toolbox, which is intended for a flexible application of VB analysis in a wide variety of scenarios in medical imaging and radiation oncology. The MAMBA toolbox is implemented in Matlab. It provides open-source functions to compute VB statistical models of the input data, according to a great variety of regression schemes, and to derive VB maps of the observed significance level, performing a non-parametric permutation inference. The toolbox allows for including VB and global outcomes, as well as an arbitrary amount of VB and global Explanatory Variables (EVs). In addition, the Matlab Parallel Computing Toolbox is exploited to take advantage of the perfect parallelizability of most workloads. The use of MAMBA was demonstrated by means of several realistic examples on a synthetic dataset mimicking a radiation oncology scenario. MAMBA is an open-source toolbox, freely available for academic and non-commercial purposes. It is designed to make state-of-the-art VB analysis accessible to research scientists without the programming resources needed to build from scratch their own software solutions. At the same time, the source code is handed out for more experienced users to complement their own tools, also customizing user-defined models. MAMBA guarantees high generality and flexibility in the design of the statistical models, significantly expanding on the features of available free tools for VB analysis. The presented toolbox aims at increasing the reach of VB studies as well as the sharing of research results.

Acceleration strategies for Tridimensional Coupled hydromechanical problems based on CPU and GPU programming in MATLAB.

Large-scale fluid flow in porous media demands intense computations and occurs in the most diverse applications, including groundwater flow and oil recovery. This article presents novel computational strategies applied to reservoir geomechanics. Advances are proposed for the efficient assembly of finite element matrices and the solution of linear systems using highly vectorized code in MATLAB. In the CPU version, element matrix assembly is performed using conventional vectorization procedures, based on two strategies: the explicit matrices, and the multidimensional products. Further assembly of the global sparse matrix is achieved using the native sparse function. For the GPU version, computation of the complete set of element matrices is performed with the same strategies as the CPU approach, using gpuArray structures and the native CUDA support provided by MATLAB Parallel Computing Toolbox. Solution of the resulting linear system in CPU and GPU versions is obtained with two strategies using a one-way approach: the native conjugate gradient solver (pcg), and the one provided by the Eigen library. A broad discussion is presented in a dedicated benchmark, where the different strategies using CPU and GPU are compared in processing time and memory requirements. These analyses present significant speedups over serial codes.

First- and second-order unconditionally stable direct discretization methods for multi-component Cahn–Hilliard system on surfaces

This paper proposes a first- and second-order unconditionally stable direct discretization method based on a surface mesh consisting of piecewise triangles and its dual-surface polygonal tessellation for solving the N-component Cahn–Hilliard system. We define the discretizations of the gradient, divergence, and Laplace–Beltrami operators on triangle surfaces. We prove that the proposed schemes, which combine a linearly stabilized splitting scheme, are unconditionally energy-stable. We also prove that our method satisfies the mass conservation. The proposed scheme is solved by the biconjugate gradient stabilized (BiCGSTAB) method, which can be straightforwardly applied to GPU-accelerated biconjugate gradient stabilized implementation by using the Matlab Parallel Computing Toolbox. Several numerical experiments are performed and confirm the accuracy, stability, and efficiency of our proposed algorithm.

Parallel spectral element method for guided wave based structural health monitoring

Parallel implementation of the spectral element method is developed in which flat shell spectral elements are utilized for spatial domain representation. The implementation is realised by using Matlab Parallel Computing Toolbox and optimized for Graphics Processing Unit (GPU) computation. In this way, considerable computation speed-up can be achieved in comparison to computation on conventional processors. The implementation includes an interpolation of wavefield on a uniform grid. The method was tested on experimental data set available on the Open Guided Waves platform. The qualitative comparison was performed on full wavefield data, whereas quantitative comparison was made directly on signals of propagating Lamb waves registered by piezoelectric transducers. In both cases, good agreement between numerical and experimental results was achieved. The proposed method is particularly useful for structural health monitoring algorithms in which signal parameters are required such as wave velocity dependence on the angle of propagation. Moreover, it enables model–assisted damage size estimation in which the damage influence curve is estimated based on large numerical data sets and sparse experimental data. Due to relatively short computation times, large data sets can be generated and used for machine learning or other soft computing methods opening up new possibilities in health monitoring of metallic and composite structures.

Parallel Computing as a Tool for Tuning the Gains of Automatic Control Laws

Select the gains of any automatic control law is not an easy task, since to do so expertise and specialized knowledge are required. Also, it demands a lot of time due to, usually, such a selection is carried out by the trial and error method, which implies to rebuild the test of the control ‑every time its gains are modified‑ until a “good” gain selection be found. Thus, this paper presents a procedure based on parallel computing, which facilitates the gain selection of an automatic control law and reduces the time spent on that. Such a procedure consists on the following four steps: i) By taking into account a tuning rule and the number of control gains, a finite set of numerical values are generated and grouped in arrays through a Matlab script. Hence, a large number of combinations to select such gains is obtained. ii) With these combinations, numerical simulations of the system in closed-loop are simultaneously performed through Matlab Parallel Computing Toolbox. iii) The several obtained system responses are treated to determine the ones achieving the control objective. iv) Lastly, the gain combination that delivers a control response with the smallest error is identified. The proposed procedure is implemented to select the gains of a state feedback control that stabilizes the Furuta pendulum in the inverted upright position. The best gain selection of the control is verified trough an experimental test with a real Furuta pendulum. The main advantages of the proposed procedure are the several gain combinations that can be simulated in a short time compared with the classical trial and error method and the effectiveness for experimental application.

An Estimation of the Learning Curve Effect on Project Scheduling with Calendar Days Calculation

Although learning is an essential part of life, traditional scheduling techniques cannot efficiently handle this effect. Considering the effects of the learning curve, it is possible to make a better prediction of project duration thus saving time and money. In this paper, a learning curve (or experience curve) effect on project duration is shown and calculated with calendar days. In the literature, project scheduling has been studied; however, only a few papers take the learning effect on project duration into consideration. The learning effect on project duration with the help of test problems and real problems was investigated. In test problems learning curve effect can occur between two consecutive activities. These pairs are chosen randomly. After calculating project duration, these pairs are allocated to be closer to each other using the predecessor's total float time. It is assumed that the duration of impending repetitive activities is shorter due to the learning curve effect if the gap between consecutive activities is small enough. This iteration is carried out until it is not possible to shorten the successor's activity time in a pair. It is shown that this effect brings a 1-3% shorter project duration. This “Calendar Days” calculation led to an integer programming problem that was solved by Matlab Parallel Computing Toolbox. 2.

A performance analysis of computing the LU and the QR matrix decompositions on the CPU and the GPU

We present an analysis of time efficiency of five different implementations of the LU and the QR decomposition of matrices performed on central processing unit (CPUs) and graphics processing units (GPUs). Three of the considered implementations, developed using the Eigen C++ library, Intel MKL, and MATLAB are executed on a multi-core CPU. The remaining two implementations are processed on a GPU and employ MATLAB's Parallel Computing Toolbox and Nvidia CUDA augmented with the cuSolver library. Computation times are compared using randomly generated single- and double-precision floating-point matrices. The experiments for the LU decomposition show that the two GPU implementations offer best performance for matrices that can fit into the GPU global memory. For larger LU decomposition problem instances, Intel MKL on the CPU is found to be the fastest approach. Furthermore, Intel MKL also proves to be the fastest method for computing QR decomposition for all considered sizes of matrices.

A parallel computing approach for integrated security assessment of power system

Abstract Fast and accurate assessment of integrated security of present day complex and highly interconnected power system within time frame is a challenging task. The traditional methods used for integrated security assessment are fast but approximate. On the other hand, full simulation methods provide more accurate security assessment, but require large computation time. This paper proposes the use of parallel computing for fast and accurate assessment of integrated security of power system by running full Newton Raphson (NR) load flow program with sparsity technique to compute real power flow or MW performance index ( MWPI ), and voltage performance index ( VPI ), and continuation power flow program with sparsity technique to compute voltage stability index ( VSI ) for all possible single line outage contingencies. The proposed work has been carried out using the Matlab Parallel Computing Toolbox Software (PCTS). To implement parallel computing, the job scheduler, a part of PCTS, divides a computationally intensive job into various tasks and distributes these tasks to its assigned workers for evaluation. Effectiveness of the proposed approach has been demonstrated on IEEE 30-bus system and a practical NREB 246-bus Indian power system. As the proposed method utilizes NR load flow and continuation load flow programs, the accurate post contingency operating status of a power system is also available, which can be used for appropriate corrective actions planning to bring the power system back into secure state. The proposed approach is quite accurate and fast and hence can be implemented for integrated security assessment of practical power systems in smart grid environment also.

ALGORITHMS AND PROGRAMS FOR STRONG GRAVITATIONAL LENSING IN KERR SPACE-TIME INCLUDING POLARIZATION

Active galactic nuclei (AGNs) and quasars are important astrophysical objects to understand. Recently, microlensing observations have constrained the size of the quasar X-ray emission region to be of the order of 10 gravitational radii of the central supermassive black hole. For distances within a few gravitational radii, light paths are strongly bent by the strong gravity field of the central black hole. If the central black hole has nonzero angular momentum (spin), a photon's polarization plane will be rotated by the gravitational Faraday effect. The observed X-ray flux and polarization will then be influenced significantly by the strong gravity field near the source. Consequently, linear gravitational lensing theory is inadequate for such extreme circumstances. We present simple algorithms computing strong lensing effects of Kerr black holes, including effects on polarization. Our algorithms are realized in a program "KERTAP" in two versions: MATLAB and Python. The key ingredients of KERTAP are: a graphic user interface, a {\it backward} ray-tracing algorithm, a polarization propagator dealing with gravitational Faraday rotation, and algorithms computing observables such as flux magnification and polarization angles. Our algorithms can be easily realized in other programming languages such as FORTRAN, C, and C++. The MATLAB version of KERTAP is parallelized using the MATLAB Parallel Computing Toolbox and the Distributed Computing Server. The Python code was sped up using Cython and supports full implementation of MPI using 'mpi4py' package. As an example, we investigate the inclination angle dependence of the observed polarization and the strong lensing magnification of AGN X-ray emission. We conclude that it is possible to perform complex numerical-relativity-related computations using interpreted languages such as MATLAB and Python.

Udkm1Dsim—A simulation toolkit for 1D ultrafast dynamics in condensed matter

The udkm1Dsim toolbox is a collection of matlab (MathWorks Inc.) classes and routines to simulate the structural dynamics and the according X-ray diffraction response in one-dimensional crystalline sample structures upon an arbitrary time-dependent external stimulus, e.g. an ultrashort laser pulse. The toolbox provides the capabilities to define arbitrary layered structures on the atomic level including a rich database of corresponding element-specific physical properties. The excitation of ultrafast dynamics is represented by an N-temperature model which is commonly applied for ultrafast optical excitations. Structural dynamics due to thermal stress are calculated by a linear-chain model of masses and springs. The resulting X-ray diffraction response is computed by dynamical X-ray theory. The udkm1Dsim toolbox is highly modular and allows for introducing user-defined results at any step in the simulation procedure. Program summaryProgram title: udkm1DsimCatalogue identifier: AERH_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AERH_v1_0.htmlProgram obtainable from: CPC Program Library, Queen’s University, Belfast, N. IrelandLicensing provisions: BSDNo. of lines in distributed program, including test data, etc.: 130221No. of bytes in distributed program, including test data, etc.: 2746036Distribution format: tar.gzProgramming language: Matlab (MathWorks Inc.).Computer: PC/Workstation.Operating system: Running Matlab installation required (tested on MS Win XP - 7, Ubuntu Linux 11.04-13.04).Has the code been vectorized or parallelized?: Parallelization for dynamical XRD computations. Number of processors used: 1–12 for Matlab Parallel Computing Toolbox; 1−∞ for Matlab Distributed Computing ToolboxRAM: Matlab’s typical RAM requirement of 196 MB is sufficient for most simulations Classification: 7.8, 7.9, 8.External routines:Optional: Matlab Parallel Computing Toolbox, Matlab Distributed Computing Toolbox Required (included in the package): mtimesx Fast Matrix Multiply for Matlab by James Tursa, xml io tools by Jaroslaw Tuszynski, textprogressbar by Paul ProteusNature of problem:Simulate the lattice dynamics of 1D crystalline sample structures due to an ultrafast excitation including thermal transport and compute the corresponding transient X-ray diffraction pattern.Solution method:The program provides an object-oriented toolbox for building arbitrary layered 1D crystalline sample structures including a rich database of element-specific parameters. The excitation, thermal transport and lattice dynamics are simulated utilizing Matlab’s ODE solver. Alternatively, the lattice dynamics can also be calculated analytically utilizing Matlab eigenproblem solver. The dynamical X-ray diffraction is computed in a parallelized matrix formalism.Restrictions:The program is restricted to 1D sample structures and is further limited to longitudinal acoustic phonon modes and symmetrical X-ray diffraction geometries.Unusual features:The program is highly modular and allows the inclusion of user-defined inputs at any time of the simulation procedure.Running time:The running time is highly dependent on the number of unit cells in the sample structure and other simulation parameters such as time span or angular grid for X-ray diffraction computations. However, the example files are computed in approx. 1–5 min each on a 8 Core Processor with 16 GB RAM available.

Spectral Clustering Algorithm: MATLAB PCT-Based Parallel Design and Implementation

Spectral clustering algorithms inevitable exist computational time and memory use problems for large-scale spectral clustering, owing to compute-intensive and data-intensive. We analyse the time complexity of constructing similarity matrix, doing eigendecomposition and performing k-means and exploiting SPMD parallel structure supported by MATLAB Parallel Computing Toolbox (PCT) to decrease eigendecomposition computational time. We propose using MATLAB Distributed Computing Server to parallel construct similarity matrix, whilst using t-nearest neighbors approach to reduce memory use. Ultimately, we present clustering time, clustering quality and clustering accuracy in the experiments.

基于JSP的Matlab远程调用和并行处理

将各种服务、软件、甚至是操作系统移植到Web上，作为互联网未来的一个发展方向，已经被很多业内人士所认同。本文设计了一套JSP网站与Matlab协同工作的系统，使用户可以通过登录网站调用通信网络仿真平台，通过前端网页输入信道参数，调用仿真程序，并通过页面将结果反馈给用户。由于仿真平台的计算量非常大，因此利用Matlab并行计算工具箱实现并行计算功能，并加入到系统中。通过程序实例验证了并行计算的正确性及优势。As one of the development directions of the Internet in the future, migrating a variety of services, software, operating system to the Web has been recognized by most people in the field. This paper developed a set of system in which the JSP Web site works together with Matlab, so that users can call the communication networks simulation platform by logging onto the site. In the system, the channel parameters are input at the front page, the simulation program is called, and the results are feedback to the user through the Web pages. Taking account of the large calculation of the simulation platform, the Matlab Parallel Computing Toolbox is used to achieve parallel computing and added to the system. The validity and advantages of parallel computing are verified by example.

Parallel Simulation of Population Balance Model-Based Particulate Processes Using Multicore CPUs and GPUs

Computer-aided modeling and simulation are a crucial step in developing, integrating, and optimizing unit operations and subsequently the entire processes in the chemical/pharmaceutical industry. This study details two methods of reducing the computational time to solve complex process models, namely, the population balance model which given the source terms can be very computationally intensive. Population balance models are also widely used to describe the time evolutions and distributions of many particulate processes, and its efficient and quick simulation would be very beneficial. The first method illustrates utilization of MATLAB's Parallel Computing Toolbox (PCT) and the second method makes use of another toolbox, JACKET, to speed up computations on the CPU and GPU, respectively. Results indicate significant reduction in computational time for the same accuracy using multicore CPUs. Many-core platforms such as GPUs are also promising towards computational time reduction for larger problems despite the limitations of lower clock speed and device memory. This lends credence to the use of highfidelity models (in place of reduced order models) for control and optimization of particulate processes.

Tensor Factorization and Clustering for the Feature Extraction Based on Tucker3 with Updating Core

Aiming at the problems of Tucker3 to large-scale tensor when applied to feature extraction, a new factorization based on Tucker3 is proposed to extract feature from the tensors. First, the large-scale tensor is divided into multiple sub-tensors so as to conveniently compute cores of sub-tensors in parallel mode with Matlab Parallel Computing Toolbox; Then, the cores of each sub-tensor are updated for reducing deviation in calculating and the similar characteristics of sub-tensors are clustered to obtain the features. Experiment results show that this methods is able to extract features rapidly and efficiently.