Advanced Materials Research Vols. 753-755

Abstract: Hypergraph partitioning is an increasingly important and widely studied research topic in parallel scientific computing. In this paper, we present a multiway hypergraph partitioning algorithm, mixed simulated annealing algorithm for global optimization and tabu search algorithm for local optimization. Experiments on the benchmark suite of several unstructured meshes show that, for 2-, 4-, 8-, 16-and 32-way partitioning, the quality of partition produced by our algorithm are on the average 6% and the maximum 17% better than those produced by partitioning software hMETIS in term of the cutsize metric.

Abstract: Considering the influence of dimensional errors,clearances,friction coefficients,external loads and flex of part comprehensively,a multi-body dynamic model of link mechanism is established by using commercial software.Assuming that the above factors follow normal probability distribution are independent with each other,a mechanism reliability analysis method by combining simulation technology and support vector machine (SVM) are proposed to reduce the computational costs. The obtained results show that the computational costs of the proposed methods are much less than the computational costs of Monte Carlo Simulation (MCS).Therefore, the proposed methods might be efficient and valuable for the reliability analysis of complex mechanism.

Abstract: Multilevel hypergraph partitioning is a significant and extensively researched problem in combinatorial optimization. In this paper, we present a multilevel hypergraph partitioning algorithm based on simulated annealing approach for global optimization. Experiments on the benchmark suite of several unstructured meshes show that, for 2-, 4-, 8-, 16-and 32-way partitioning, although more running time was demanded, the quality of partition produced by our algorithm are on the average 14% and the maximum 22% better than those produced by partitioning software hMETIS in term of the SOED metric.

Abstract: In recent years, the highly parallel graphics processing unit (GPU) is rapidly gaining maturity as a powerful engine for high performance computer. However, in most computational fluid dynamics (CFD) simulations, the computational capacity of CPU was ignored. In this paper, we propose a hybrid parallel programming model to utilize the computational capacity of both CPU and GPU. Considering the memory amount of CPU and GPU, we also propose an out-of-core method to increase the simulation scale on single node. The experiment results show that the programming model can utilize the computational capacity of both CPU and GPU efficiently and the out-of-core method can increase the simulation scale on single node.

Abstract: Pipeline transportation is the most important transport mode for crude oil. When oil pipeline operates at low throughput or shutdown, congelation accident always happen as the temperature of oil in pipeline drops, which causes malignant accidents and significant financial losses. Study of the various factors affecting congelation failure is necessary. According to the statistics and analysis of congelation accident, direct and indirect reasons of affecting pipeline congelation failure were determined. Combining with fuzzy comprehensive assessment method, the related importance of factors on congelation accident was studied, and the grade of factors affecting failure congelation of pipeline was also identified. Aiming at the factors, some improvement measures were put forward. These can provide evidences for accident treatment, planned maintenance, security operation and scientific management.

Abstract: Product design, energy consumption and carbon emission belong to different domain. For the purpose of achieving low carbon and energy conservation design, the relation between design information and energy information must be established. This paper employs bond graph theory for extracting and expressing basic energy characteristics and presents a mapping framework between design information and energy information.

Abstract: This paper studies the scheduling problem of Hybrid Flow Shop (HFS) under the objective of minimizing makespan. The corresponding scheduling simulation system is developed in details, which employed a new encoding method so as to guarantee the validity of chromosomes and the convenience of calculation. The corresponding crossover and mutation operators are proposed for optimum sequencing. The simulation results show that the adaptive Genetic Algorithm (GA) is an effective and efficient method for solving HFS Problems.

Abstract: In this study, the author focus on the exchange rate forecasting. Exchange rates fluctuation is extremely complex, not only contains the linear part but also includes non-linear elements, In this paper, Simulated Annealing Algorithm is introduced to overcome the neural network easy fall into local minimum defects in BP neural network basis, in order to optimize the network weights and thresholds, and thus improve the prediction accuracy. Through several forecast experiments about the major currencies against, the result show that compare to the single use of BP neural network, after introduced Simulated Annealing Algorithm, the prediction accuracy and stability has been further improved, meanwhile time-consuming less than genetic algorithms and other optimization algorithms.

Abstract: The weight is description of the risk intensity. The paper using special equipment user as object to researched the equipment risk evaluation indicator system and weights. Firstly, according to the nature of enterprises, the special equipment user has been divided into four categories: the user of passenger ropeway and large recreational facilities; the user of pressure equipments; the manufacture which the pressure equipments are not the main production equipments; the user of non manufacture which are not the main production equipments. Secondly the AHP has been applied to calculate the basic weight of special equipment user risk evaluation indicator system, and the weight index has been applied to adjustment the weights. Finally the SINOPEC CATALYST company has been applied to verify the model.

Abstract: Seeing that human face features are unique, an increasing number of face recognition algorithms on existing ATM are proposed. Since face detection is a primary link of face recognition, our system adopts AdaBoost algorithm which is based on face detection. Experiment results demonstrated that the computing time of face detection using this algorithm is about 70ms, and the single and multiple human faces can be effectively measured under well environment, which meets the demand of the system.