Hierarchical optimization algorithm-based solutions to a class of bilevel programming problems

A decomposition-coordination-based optimization algorithm in a two-level hierarchical structure is proposed to solve a bilevel programming problem. By introducing decoupling variables, or coordinating variables, the programming problem is decomposed into some independent subproblems which are easily solved at Level 1 of the structure(.) And at Level 2 the coordinating variables are updated to improve the solutions to subproblems. The algorithm is carried out in an iterative fashion in order to continuously coordinate the solutions towards the optimal one to the bilevel programming problem. A case study demonstrates its feasibility and effectiveness.