An NE/SQP method for the bounded nonlinear complementarity problem
IL (United States) Argonne National Lab., S Gabriel
Abstract
IL (United States) Argonne National Lab., S Gabriel
Abstract
NE/SQP is a recent algorithm that has proven quite effective for solving the pure and mixed forms of the nonlinear complementarity problem (NCP). NE/SQP is robust in the sense that its direction-finding subproblems are always solvable; in addition, the convergence rate of this method is Q-quadratic. In this paper the author considers a generalized version of NE/SQP proposed by Pang and Qi, that is suitable for the bounded NCP. The author extends their work by demonstrating a stronger convergence result and then tests a proposed method on several numerical problems.
A significance statement is not available in the OpenAlex record.
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.
NE/SQP is a recent algorithm that has proven quite effective for solving the pure and mixed forms of the nonlinear complementarity problem (NCP). NE/SQP is robust in the sense that its direction-finding subproblems are always solvable; in addition, the convergence rate of this method is Q-quadratic. In this paper the author considers a generalized version of NE/SQP proposed by Pang and Qi, that is suitable for the bounded NCP. The author extends their work by demonstrating a stronger convergence result and then tests a proposed method on several numerical problems.
Key concepts: Sequential quadratic programming, Nonlinear complementarity problem, Complementarity (molecular biology), Mixed complementarity problem, Bounded function, Mathematical optimization, Mathematics, Complementarity theory