Dissertation > Excellent graduate degree dissertation topics show

Social emotional optimization algorithm in the cluster structure optimization

Author: ChenYongJing
Tutor: ZengJianChao；CuiZhiHua
School: Taiyuan University of Science and Technology
Course: Applied Computer Technology
Keywords: Global Optimization Social emotional optimization algorithm L-BFGS The LJ atom clusters and Ag clusters Seed Technology Grid search technology
CLC: TP301.6
Type: Master's thesis
Year: 2011
Downloads: 26
Quote: 0
Read: Download Dissertation

Abstract


Cluster structure of the optimization problem is a NP problem, the main difficulty lies in the number of local minima grows exponentially with the increase of the number of atoms, and the local minima and the global minimum is very close, the algorithm easy to fall into local minima. LJ issues and Ag clusters structural problems are two typical cluster optimization problem. LJ problem of application is very broad, the LJ function can be used not only to structural optimization of the inert gas and the description of the protein folding problem, and can be used to study the interaction between the C atom and Cs atoms and carbon pods carbon nano carbon nanotubes C-C interactions between the tube and spherical fullerenes. Efficiently solve the the LJ structural problems of development of major significance to the field of molecular biology and materials science. Ag clusters have important applications in catalysis, electronic materials and alloys. Social emotional optimization algorithm is a new swarm intelligence optimization algorithm. Algorithm, each individual represents a virtual, in the virtual society through cooperation, competition and other ways to improve their own social status which. Each the evolution algebra (time fragment) they will make decisions based on the the corresponding sentiment index, their own experience and the social experience and choose what they think can improve their own social status, the result will be the degree of social recognition feedback the behavior correctness. If the behavior is recognized, the sentiment index rose, and otherwise reduce. Social-emotional optimization algorithm for cluster structures optimized the Ag cluster structures of the LJ atom clusters and atomic number atomic number less than 17 is less than 9. Optimization is very effective structure of the seed clusters, using seed technology optimization allows the algorithm to guide the search to improve search efficiency, the disadvantage is that the seed selection difficult to select the cluster structures of similar configuration as seed atomic distribution of selected seeds with different or outer layer of the structure to be optimized configuration of clusters vary widely, it will make the algorithm into a local optimum. In order to improve the efficiency of the optimization of the algorithm, this paper introduces seed technology, reducing the number of optimization optimize difficulty LJ atomic clusters of 310 Ag cluster structures optimized Ag43 structure. The grid search technique is optimized atom is placed in a location where it may exist, and optimized by the algorithm to select an optimal location in a location where it may exist. Social-emotional optimization algorithm discretization introduction of grid point search technology, reduce the scope of the search, the search space into several discrete points, its optimization results significantly better than the first two optimization methods. Shorten the optimization time of the algorithm in order to simplify the algorithm, based on the growth characteristics of the cluster structure ---- atom clusters from the inside outward growth, will be arranged in the outer layer of the inner layer of atoms lined atom clusters in this article optimization only to optimize the the outermost layer structure, the the default atomic structure of the inner layer is filled shells. This will not only improve the success rate of the algorithm, but also shorten the optimization time. And optimization of LJ582.

Related Dissertations

  1. Global Optimization for Sum-of-Ratios Problems,O224
  2. Research on the Peak to Average Power Ratio Reduction for OFDM System by BFGS,TN919.3
  3. Structural constraints of the dynamic PET image reconstruction,TP391.41
  4. The Improvement and Research of Several Algorithms about the Global Optimization,O224
  5. The Research and Application on Composition Papers System Base on Genetic Algorithm,O224
  6. The Auxiliary Function Method for Nonlinear Global Optimization,O224
  7. Spectral-scaling Quasi-newton Method for Solving Nonlinear Monotone Equations,O241.6
  8. A New Class of Quasi-newton Algorithm and Its Convergence,O224
  9. Globally Convex Filled Function Method for Nonlinear Global Optimization Problem,O221.2
  10. Transmission Network Expansion Planning Based on Ecology Evolutionary Algorithm of Food Chain,TM715
  11. Multi-stream heat exchanger channel arranged to optimize the design of,TK172
  12. The Study of Global Optimization Hybrid Intelligent Algorithm Based on Particle Swarm Optimization,TP18
  13. Calibration of Local Volatility Surfaces Using Tikhonov Regularization,F830.9
  14. A Study on the Multi-Population Evolutionary Algorithms for Discrete Variables,TP18
  15. Research on Node Localization Technology Based on RSSI in Wireless Sensor Networks,TP212.9
  16. Research of Three-Dimensional Localization Algorithm for Wireless Sensor Networks,TN929.5
  17. Study on the Technology for Mosaic of the Power Semiconductor Device’s Image,TN303
  18. Based on support vector machine and orthogonal wavelet transform blind equalization algorithm,TN911.5
  19. Global Optimization Algorithm Based on Visual Cognition,TP301.6
  20. Deformation monitoring data processing model based on wavelet theory research,TU196.1
  21. Based upon digestion items identified global optimization method coreference resolution,TP391.1

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > General issues > Theories, methods > Algorithm Theory
© 2012 www.DissertationTopic.Net  Mobile