Filled Function Algorithm For Nonlinearly Global Optimization System

As a mathematical method， global optimization problem has been a very widerange of applications in real life, especially in engineering, molecular biology, neuralnetworks, and social sciences, so global optimization problems has attracted theattention of many researchers. However, due to multiple extreme points of globaloptimization problem, the researchers have to face the following problems: One is howto get another from a minimum point smaller than the minimum point, and the other ishow to determine the current minimum point is a global minimum point. Thesequestions can not be used the traditional nonlinear programming problem to solve.There are many algorithms to solve global optimization problem, and these algorithmscan be roughly divided into two categories, namely deterministic algorithm andrandomness algorithm, in which the filled function method is a deterministic algorithm.In this article, we mainly do research on nonlinear optimization problem of the largesystem.We mainly study from the following aspects:The first chapter gives a brief introduction to optimization problems in life, givessome specific concepts about optimization problems, and also introduces severalmethods to optimization problems. Finally，describe the history of the development andthe basic methods of the fill function.The second chapter, according to the way of thinking and the basic theory of thefilled function algorithm, we proposed a one-parameter filled function for solvingglobal optimal solution, give the algorithm, analysis and discussion its excellentproperties.The third chapter, filled function given by the predecessors contains one or twoparameters. More complex due to the selection of parameters, on the basis of theprevious, the chapter gives a filled function with no parameters to solve a class ofnon-smooth global optimization, discusses the algorithm and its excellent properties andgives numerical experiments.The firth chapter，brief introduction to the basic concepts and the nature of the discrete filled function, give a class filled function for solving integer programmingproblems, prove its excellent properties, illustrate the effectiveness of the algorithm bynumerical experiments.The fifth chapter, summary and outlook filled function algorithm development.