Advanced Materials Research Vols. 694-697

Abstract: For the multiple solutions of wire tensions to wire-driven parallel manipulator, the paper presents a wire tension optimization program based on the penalty function. After building a nonlinear programming model and simulating in MATLAB, the optimal solution of wire tensions under the minimal preload is calculated. Compared with the simulation results and experimental results, it shows that the wire tensions in experiment are relatively stable and the trends are overall consistent with the simulation results. The study lays a foundation for selecting the suitable preload for the wire-driven parallel manipulator.

Abstract: The subproblem of the globally convergent version of method of moving asymptotes (GCMMA) was deeply studied and realized by Lagrange dual method, sequential unconstrained minimization technique (SUMT) and conjugate gradient method. Based on the convex, separable and conservative properties of subproblem, a dual problem was built by using Lagrange dual method. The dual problem was transformed into unconstrained optimization problem using SUMT and solved by conjugate gradient method. Finally, the feasibility of solution of subproblem is demonstrated by truss structural optimization problem. Comparing with other optimization algorithms, GCMMA can converge to the global optimal solution.

Abstract: Many complex discrete manufacturing problems, such as manufacturing sequencing problem or machine scheduling problem etc, can be converted into a general traveling salesman problem (TSP). TSP has been proven to be NP-complete. The genetic algorithm is improved with two heuristic rules for TSP. The first heuristic rule is the four vertices and three lines inequality. It is applied to the local Hamiltonian paths to generate the better solutions. The second heuristic rule is executed to reverse the local Hamiltonian paths, which generates new better solutions. The two heuristic rules coordinate with each other and they are merged into the optimization process of genetic algorithm to improve its performance. The computation results show that the improved genetic algorithm can find the near optimal solutions for most of the TSP instances.

Abstract: Clustering analysis, as a practical data mining method, has wide-ranging applications in many fields. But because of different original data resources, clustering results of different data distribution patterns and applicable clustering evaluation methods are different from each other. Aiming at different data distribution patterns, only reasonable clustering evaluation methods can achieve a better recognition of different clustering results for realizing the application value of clustering technology. In this paper, the combination clustering evaluation model is constructed form three angles, through clustering experiment of different artificial simulated data distribution patterns, comparative analysis draw a conclusion that the combination clustering evaluation model constructed is reasonable, and according to applicable clustering and clustering evaluation methods based on different data distribution patterns, the optimization clustering process is constructed for improving the effectiveness and interpretability of different clustering results.

Abstract: In this paper, we obtained the existence of periodic solutions for a class of even order delay differential equations. It is available in the Vibration, Noise Analysis and Control. Based on combining a method of Fourier analysis and the Use of the Schauder fixed point theorem, we give the proofs. This generalizes results developed by W. Layton to higher order equations.

Abstract: The purpose of this paper is to present some results for Lagrange-Fenchel dualities, via a perturbation approach and a conjugate mapping for the set-valued optimization which is under the assumption of Benson efficiency.

Abstract: As an important method for secret shared and key recovery since has been proposed, threshold scheme has been extensively studied and applied. In this paper, an improved threshold scheme based on the order is proposed in order to meet actual application's demands and enhances the security of the traditional threshold scheme. This new scheme can approximate immensely expand the number of permutation so that even the attacker gets all secret shares but difficult to recover the original secret. So the new scheme proposed in this paper can increase the security of the threshold scheme greatly and expands the application occasions of the threshold scheme.

Abstract: This project designed and accomplished a series of aerodynamics virtual experiment software including To Mensurate Airfoil Pressure Distribution based on Computational Fluid Dynamics, multimedia technologies, and other modern computer technologies. The results showed that the series of virtual experiment software are successful and effectual. It was concluded that virtual experiment designed scientifically and rationally can reform practical teaching.

Abstract: It is well known that an engineering surface is composed of a large number of wavelengths of roughness that are superimposed on each other. Because these multi-scale features are related to different aspects of the processes the surface has undergone and closely related to the friction and wear properties of a surface, the analysis and characterization of these features becomes an important aspect of manufacture. The challenge is how to use them for acquiring knowledge and for aid to analysis. In this paper, a method for surface topography analysis is proposed based on bidimensional empirical mode decomposition (BEMD), which can provide good adaptive separation of surface texture into multiple hierarchical components known as bidimensional intrinsic mode functions (BIMFs). Applications are conducted by using a simulated surfaces to demonstrate the feasibility and applicability of using the bidimensional empirical mode decomposition method in the analysis of engineering surfaces.

Abstract: To evaluate the yarn tension detection and control schemes in rapier looms, a fuzzy multiple-attribute group decision making problem is proposed for the schemes selection. Firstly, important degrees of every attributes from each expert are considered. The individual opinions of each expert are integrated with the similarity of the decision group. And the synthesized weights of each expert are calculated. Secondly, with the aggregation of experts opinions, the group attribute-weights matrixes are obtained. Then the fuzzy TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) is used to sequence the alternatives, and the optimal scheme is decided for yarn tension detection and control system, the decision results illustrate the feasibility and effectiveness of the developed method.