Hierarchical optimization method for a class of nonlinear bilevel programming problems
Ruibo Li
Abstract
Ruibo Li
Abstract
A novel method for a class of nonlinear bilevel programming problems is proposed. By introducing a (decoupling) vector, a bilevel programming problem is decomposed into independent optimization sub-problems, which are easily solved at level 1 of a two-level hierarchical structure. At level 2 the decoupling vector is then updated (using) solutions from level 1. Based on decomposition-coordination principle, the proposed method can finally solve the optimal solution of the bilevel programming problem in an iterative fashion. For programming problems with (integers,) continualization technique is employed and continualized problems can be easily solved using the proposed method. Numerical examples are used to demonstrate simplicity and effectiveness of the proposed method.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A novel method for a class of nonlinear bilevel programming problems is proposed. By introducing a (decoupling) vector, a bilevel programming problem is decomposed into independent optimization sub-problems, which are easily solved at level 1 of a two-level hierarchical structure. At level 2 the decoupling vector is then updated (using) solutions from level 1. Based on decomposition-coordination principle, the proposed method can finally solve the optimal solution of the bilevel programming problem in an iterative fashion. For programming problems with (integers,) continualization technique is employed and continualized problems can be easily solved using the proposed method. Numerical examples are used to demonstrate simplicity and effectiveness of the proposed method.
Key concepts: Bilevel optimization, Decoupling (probability), Mathematical optimization, Nonlinear programming, Nonlinear system, Single level, Computer science, Class (philosophy)