| Along with human society's development, economical globalization aggravating, in many practical problems, decision problem is completed by many decision-makers, and these decision-makers have several levels and relative independence, so the level feature of decision problem becomes more and more obvious. Hence, the research of bilevel programming problems has vital practical significance and applied prospect. However, it is very difficult to determine its solution because of its inherent nonconvexity and nondifferentiability. In particular, it is more difficult to get the global optimal solution of nonlinear bilevel programming. On the other hand, because many difficult nonlinear programs usually can be solved by Sequential Quadratic Programming Method, moreover, during the past several decades, quadratic programming already became essential methods for operations research, economical mathematics, management science, system analysis and combination optimization and so on, quadratic programming has received increasing attention of specialists and scholars. This paper will combine quadratic programming and bilevel programming to do further analyzes and discussion.Firstly, the nature, principle and model of bilevel programming, quadratic bilevel programming in particular, are analyzed and discussed in this paper. And based on these analyzes and discussion, effective decomposition and transformation of quadratic bilevel programming are proposed from different aspects. Secondly, this paper has given some important optimality conditions for quadratic bilevel programming in more ordinary case. Finally, a new solution method for quadratic programming and two new solution methods for quadratic bilevel programming are proposed by combining the newest research achievements of quadratic programming and bilevel programming. |
