Dissertation > Excellent graduate degree dissertation topics show

Improved evolutionary algorithm based on the entropy of the amount of conserved

Author: DaiZhiHuang
Tutor: LiKangShun
School: Jiangxi University of Technology
Course: Applied Computer Technology
Keywords: Evolutionary Algorithm Entropy conservation Non - uniform mutation rate Semi- consistent crossover operator Function optimization
CLC: TP301.6
Type: Master's thesis
Year: 2010
Downloads: 53
Quote: 1
Read: Download Dissertation

Abstract


This thesis is composed of three parts, the first part is the introduction of the research background, given the course of evolutionary algorithm development, research status and development prospects; brief analysis of the second part of the evolutionary algorithm, evolutionary algorithm is introduced some of the basic concepts, key elements of the evolutionary algorithm, some improvements mathematical evolutionary algorithm and evolutionary algorithm theoretical basis; third part of the first of two parts on the basis of the traditional evolutionary algorithm improvements proposed based on the entropy conservation Improved Evolutionary Algorithm (An Improved Evolutionary Algorithm Based on Law of Conservation of Entropy, hereafter ECEA algorithm); fourth part is the part of the specific application of the evolutionary algorithm: the proposed ECEA algorithm used in typical with constraints complex function optimization problems, the results were compared with the results with other algorithm, the results show that our proposed algorithm has better convergence rate and solution accuracy. The paper work and innovation: (1) based on the conservation of entropy amount Improved Evolutionary algorithm, which draw on the law of conservation of the amount of entropy in the cosmic system to coordinate the contradiction between the diversity of the population and the elite strategy. So elite select adaptive selection based on the diversity of the population. At the initial stage of the iterative algorithm, the search area by reducing the choice of solutions for elite increases Solutions. With the the iterative algebra increase, dynamically increase the number of selected elite solution to accelerate the speed of convergence of the algorithm. (2) proposed a semi-consistent crossover operator, it can be based on changes in the diversity of the population distribution adaptive crossover operator. Before the mid-term increase in the diversity of the population so that the population evolution, the latter enables the population to accelerate convergence to the optimal value. Based on population entropy change in non-uniform mutation rate, its population evolution before the mid-term value is relatively small, thus contributing to the excellent individual is not destroyed, while in the later stage of evolution values ??have increased, which to a certain extent reduce the populations of the possibility of convergence to a local optimal solution. (3) through the theoretical proof and instance comparison to illustrate the feasibility and validity of ECEA. First, according to the theory of finite Markov chain the ECEA convergence analysis and algorithm computing performance. The ECEA algorithm is then applied to some Benchmark with constrained optimization problem of complex functions, and compared with other literature, the results show that the algorithm converges fast, high-precision.

Related Dissertations

  1. Evolutionary Clustering Algorithm and Its Application,TP311.13
  2. Knowledge Evoluation Algorithm and Its Application in Chemical Dynamic Optimization,TQ021.8
  3. Quantum Clonal Multiobjective Evolutionary Algorithms,O413
  4. Optimization of the process parameters of the cigarette filter rod forming equipment opener,TS43
  5. Research on 10kV Distribution Network Reconfiguration Based Evolution Algorithm,TM715
  6. Research on Improved Differential Evolutionary Algorithm for Multi-Objective Optimization,TP301.6
  7. SAR Image Segmentation Based on Evolutionary Computation,TN957.52
  8. Particle swarm optimization algorithm and its application in nonlinear regression models,O212.1
  9. The Research of a Multi-objective Evolutionary Algorithm Based on Steady-state,O221.6
  10. Researches on Fireworks Explosion Optimization Algorithm and Its Improvement,O224
  11. Research on Optimal Control for the Fermentation Process,TQ922.1
  12. Transmission Network Expansion Planning Based on Ecology Evolutionary Algorithm of Food Chain,TM715
  13. Chaotic Monkey Algorithm and Its Applications,TP18
  14. Research on MOEA in Searching Robust Optimal Solutions,O224
  15. The Charateristic of Time-Frequency Atom Dictionary Research on Radar Emitter Signal,TN974
  16. The Grid Mechanism for Multi-Objective Evolutionary Algorithm,TP301.6
  17. Quantum-Inspired Evolutionary Algorithm and Its Application in Image Sparse Decomposition,O413
  18. Research and Application on Constratined Multi-objective Evolutionary Algorithm,O224
  19. Research and Application on Evolutionary Multi-Objective Algorithm Based on Arena Principle and Niche,O221.6
  20. Agent Based Evolutionary Algorithms Applied to Constrained Multi-Objective Optimization and Decision-Making,TP18
  21. Evolutionary Strategies for Solving Dynamic Optimization Problems,TP18

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