the permutation flow shop sequencing problem withthe objective of minimizing the sum of the job's completion times, in literature known as the Fparallel toC(sum), has been considered. the parallel genetic algorit...
详细信息
ISBN:
(纸本)3540219463
the permutation flow shop sequencing problem withthe objective of minimizing the sum of the job's completion times, in literature known as the Fparallel toC(sum), has been considered. the parallel genetic algorithm based on the island model of migration has been presented. By computer simulations on Taillard benchmarks [10] and the best known results from literature [9] we have obtained not only acceleration of the computation's time but also better quality and stability of the results.(1).
In the paper a new parallel method for learning decision rules is proposed. the method uses evolutionary algorithm to discover decision rules from datasets. We describe a parallelization of the algorithm based on mast...
详细信息
ISBN:
(纸本)3540219463
In the paper a new parallel method for learning decision rules is proposed. the method uses evolutionary algorithm to discover decision rules from datasets. We describe a parallelization of the algorithm based on master-slave model. In our approach the dataset is distributed among slave processors of a parallel system. the slave procesors compute fitness function of chromosomes in parallel. the remainder of evolutionary algorithm i.e. selection and genetic search operators is executed by the master processor. Our method was implemented on a cluster of SMP machines connected by Fast Ethernet. the experimental results show, that for large datasets it is possible to obtain a significant speedup.
Peer-to-peer (P2P) computing has emerged as an alternative model of communication and computation to client-server model. While, P2P computing may significantly increase the performance and the scalability of the whol...
详细信息
ISBN:
(纸本)3540219463
Peer-to-peer (P2P) computing has emerged as an alternative model of communication and computation to client-server model. While, P2P computing may significantly increase the performance and the scalability of the whole system, they still are facing many challenges in achieving these goals. In this paper we study the problem of scheduling a large number of tasks on such systems. We propose two techniques based on heuristic approach: a Recursive Neighbour Search and an Augmented Tabu-Search technique. these techniques do not address directly the load-balancing problem since it is completely unrealistic in such large environments, but we will show that even a non-perfectly load-balanced system can behave reasonably well by taking into account the tasks' time demands. these algorithms are compared to a well known scheduling algorithm in order validate their performance.
this paper presents the simulation of 2D drywet radial dambreak cases using shallow water equations on multicore processing. Here, OpenMP for parallel computing is used. the CPU time for parallel programming with N = ...
详细信息
ISBN:
(纸本)9781509049127
this paper presents the simulation of 2D drywet radial dambreak cases using shallow water equations on multicore processing. Here, OpenMP for parallel computing is used. the CPU time for parallel programming with N = 800 points is shown almost similar to the the CPU time for serial programming with N = 400. the speedup and efficiency using parallel with 16 processors are obtained 7.8 times for dry simulation, 8.6 times for wet simulation and 50% from the serial respectively. this means the computational cost using parallel reduces 50% from the average of computational cost by a single processor.
In this paper we develop a high-level parallel program synchronization and control mechanism. A specialized controller process (or a hierarchy of such processes) collects information about process states and construct...
详细信息
ISBN:
(纸本)3540219463
In this paper we develop a high-level parallel program synchronization and control mechanism. A specialized controller process (or a hierarchy of such processes) collects information about process states and construct Strongly Consistent Global States using time interval timestamps. Global predicates evaluated by controllers prompt sending of control signals to processes and the signals trigger computation activation or cancellation. the proposed framework is integrated with a message passing system - it is added to the GRADE visual parallel programming environment to enhance its message-passing based features. the GRADE enhancements introduced here are discussed.
this paper presents parallel strategies applied to Harmony Search ( HS) on CUDA for solving Traveling Salesman Problem ( TSP), which is an NP-hard problem in combinatorial optimization. the implementations are tested ...
详细信息
ISBN:
(纸本)9781467365376
this paper presents parallel strategies applied to Harmony Search ( HS) on CUDA for solving Traveling Salesman Problem ( TSP), which is an NP-hard problem in combinatorial optimization. the implementations are tested with data instances of varying problem size from TSPLIB to measure speedup attained over serial implementation. In our experiment, HSGPU parallel implementation can acquire 32 times speedup as compared with CPU- based implementation.
In Niederreiter's factorization algorithm for univariate polynomials over finite fields, the factorization problem is reduced to solving a linear system over the finite field in question, and the solutions are use...
详细信息
ISBN:
(纸本)3540219463
In Niederreiter's factorization algorithm for univariate polynomials over finite fields, the factorization problem is reduced to solving a linear system over the finite field in question, and the solutions are used to produce the complete factorization of the polynomial into irreducibles. For fields of characteristic 2, a polynomial time algorithm for extracting the factors using the solutions of the linear system was developed by Gottfert, who showed that it is sufficient to use only a basis for the solution set. In this paper, we develop a new BSP parallel algorithm based on the approach of Gottfert over the binary field, one that improves upon the complexity and performance of the original algorithm for polynomials over F-2. We report on our implementation of the parallel algorithm and establish how it achieves very good efficiencies for many of the case studies.
In the aircraft industry structural components, referred to as part numbers (PN), have to be subject to an heat treatment in capacitated burn-in furnaces for a pre-defined period (exposure time) in order to provide th...
详细信息
ISBN:
(纸本)9781467358149;9781467358125
In the aircraft industry structural components, referred to as part numbers (PN), have to be subject to an heat treatment in capacitated burn-in furnaces for a pre-defined period (exposure time) in order to provide them with specific physic and chemical features (e.g. hardness, corrosion resistance, conductivity). Two or more part numbers can be grouped in a batch and treated simultaneously in the same furnace if it is possible to individuate a common exposure time. In order to minimize the total completion time (makespan) of the process it needs to determine the appropriate grouping of the part numbers into batches (batching problem) to be processed by each furnace (scheduling problem). the problem can be modeled as a batch scheduling problem on parallel machines where the batching and the scheduling problem are considered at the same time. Starting from a real case study, we present an original integer linear programming formulation in the case of two capacitated parallel machines and we provide the results obtained on two real instances coming from the aircraft industry.
the aim of this article is to describe the design and the implementation of skeletons for the Branch-and-Bound technique. the library developed provides sequential and parallel resolution patterns to solve optimizatio...
详细信息
ISBN:
(纸本)3540219463
the aim of this article is to describe the design and the implementation of skeletons for the Branch-and-Bound technique. the library developed provides sequential and parallel resolution patterns to solve optimization problems through this technique on a wide range of parallel computing platforms. To use the skeletons proposed to solve a particular problem, the skeleton filler has to specify the type of the problem, the type of the solution and the specific characteristics of this paradigm using the C++ programming language, and the user does not need to have any knowledge in parallelizations techniques. this paper describes the sequential and the parallel algorithms for message passing and shared memory platforms, and shows computational results for the 0-1 knapsack problem on an Origin 3000.
In this paper we present a software for parallel solution of engineering problems based on the boundary element method. the library is written in C++ and utilizes OpenMP and MPI for parallelization in both shared and ...
详细信息
In this paper we present a software for parallel solution of engineering problems based on the boundary element method. the library is written in C++ and utilizes OpenMP and MPI for parallelization in both shared and distributed memory. We give an overview of the structure of the library and present numerical results related to 3D sound-hard scattering in an unbounded domain represented by the boundary value problem for the Helmholtz equation. Scalability results for the assembly of system matrices sparsified by the adaptive cross approximation are also presented. (C) 2016 international Association for mathematics and Computers in Simulation (IMACS). Published by Elsevier B.V. All rights reserved.
暂无评论