Heuristic optimization algorithm based on Cluster Analysis

Abstract

 

This article covers optimization algorithms (Classic optimization algorithms and heuristic optimization algorithms), algorithm complexity theory, clustering analysis, and other related fields. A more detailed summary of heuristic optimization algorithms, and a general summary of computing complexity theory and clustering analysis. clustering Analysis and group sequence are introduced in the process of solving the optimization problem. The concept of the Set descent direction is introduced, and a calculation method of the Set descent direction is given. the descent direction of the Set improves the search efficiency of General heuristic algorithms, so that they can calculate a better effective solution within a limited time. Then, based on the descent direction of the set, the clusteropt algorithm is proposed and compared with the genetic algorithm by using a numerical example (Multi-Peak Function) to verify its effectiveness.

 

The first chapter provides the research background and objectives of the clusteropt algorithm. optimization Algorithms (mainly classical optimization algorithms and heuristic optimization algorithms), algorithm complexity theory, clustering analysis, and other related fields. A more detailed summary of the heuristic optimization algorithm, and a general summary of the computing complexity theory and clustering analysis; the second chapter introduces the idea of clustering analysis and group sequence in solving the optimization problem, the concept of a set descent direction is proposed, and the calculation method of the Set descent direction is given. the descent direction of the Set improves the search efficiency of General heuristic algorithms, so that they can calculate a better effective solution within a limited time; chapter 3 describes the algorithm structure of the clusteropt algorithm and some numerical techniques involved in the actual computation. Chapter 4 describes the numerical test of the clusteropt algorithm. For the multi-Peak Function example, compare the cluster analysis-based clusteropt algorithm with the genetic optimization algorithm (GA) to verify its effectiveness and fast convergence. Chapter 5 summarizes and prospects, and analyzes the characteristics of the clusteropt algorithm, and put forward some suggestions for improvement.

 

Key words: heuristic algorithm clustering analysis set descent direction group order

Thesis PDF

Http://ariszheng.googlepages.com/---V1.0.doc.pdf