Some Researches On The Filled Function Methods For Global Optimization Problems

Global optimization problems found in a wide range of engineering, military,national defense, economy and many other fields. The existing solution for nonlinearprogramming problems with most methods are only calculated the local minimumproblem. However, in scientific research and engineering design, involvingoptimization problems need to find out the global optimal solution, but the problemitself often has many different from the global minimum local minimum, so that theglobal optimization problems are usually not with general nonlinear programmingmethod to obtain the global optimal solution.The filled function method is to use more mature local optimization method toobtain the global minimum point. But, at present the exist filled function method intheoretical or computational have different degrees of defects. How to improve thefilled function method in the theory of defects, and how to design more effective filledfunction algorithm is very important. In view of this, this article on the filled functionmethod is further discussed.About filling function method has a lot of research and results. In this paper, themain consideration unconstrained optimization problems, mainly in the previous workfoundation, tried to give a new single parameter filling function and an solving integerprogramming single parameter filling function, analysis and prove their severalproperties, thus design new filling function algorithm.This paper structure is divided into four parts, the first part of this globaloptimization and reserve knowledge, in second part LiPuXi without this continuouscondition, to the general unconstrained put forward a new simple single parameterfilling function, analysis and proves the filling of the nature of the function. The thirdpart is to solve the problem of integer programming is proposed as a new single parameter filling function. The fourth part is the conclusion and prospect.