Solving Nonlinear Programming Problem Based On Genetic Algorithm

Genetic Algorithm is one of the current hot topics in the area of computer science. The massive parallelism, generality and flexibility it contains have attracted much attention from lots of experts in different fields. Solving nonlinear programming problem with genetic algorithm is the most important director of the application of genetic algorithm.The main content of the thesis is discussing the genetic algorithm and designing genetic algorithms for solving nonlinear programming problems. In addition, we propose reliable, efficient algorithms for solving problems.First, this thesis gives a brief introduction to the feature and application fields of the nonlinear programming and some methods often used to solve nonlinear programming problems. Several success solutions based on genetic algorithm will also introduced.In the second chapter, the basic concepts and theory of genetic algorithms are introduced first and then the design of genetic algorithm has been discussed, some classical strategies used by genetic algorithm to solve nonlinear programming problems will be shown in this chapter as well.In the third chapter, We also propose some methods to optimize the parameters of the hybrid genetic algorithm, when used these ways in hybrid genetic algorithm (GA+Simplex), the results we got are much better than before.In the fourth chapter, several methods for handling constrains by genetic algorithms for nonlinear programming problems are introduced. These methods were grouped into four categories: methods based on preserving feasibility of solutions, methods based on penalty functions, methods which make a clear distinction between feasible and infeasible solutions and other hybrid methods. We analyze such methods in detail and designed a new penalty function, experiment result proved that the genetic algorithm used such penalty function can get better quality of the solution in solving nonlinear programming problems.At the end of the chapter, the software complied based on the methods introduced in beyond chapters are introduced. Using this software, we can change the genetic operators as we want and add new operators are very convenience. Results compare this software and some commercial or noncommercial optimization software are listed in the end of the chapter.