Title :
A Modified Penalty-Free-Type Nonmonotone SQP Method
Author_Institution :
Coll. of Math. & Comput. Sci., Hebei Univ., Baoding, China
Abstract :
A modified penalty-free-type nonmonotone SQP method for inequality constrained optimization is proposed in this paper. This algorithm has no need on the penalty function and allows nonmonotonic decrease for the constrained violation function. Comparing with the filter-based approaches, our method leads to more flexibility. Under some reasonable conditions, the global convergence properties of our algorithm is presented. Moreover, some numerical results are given.
Keywords :
constraint theory; quadratic programming; constrained violation function; filter based approach; flexibility; global convergence property; inequality constrained optimization; modified penalty free type nonmonotone SQP method; penalty function; Convergence; Indexes; Newton method; Programming; Quadratic programming; Robustness; Global convergence; Nonmonotone; Sequential quadratic programming;
Conference_Titel :
Computational Sciences and Optimization (CSO), 2011 Fourth International Joint Conference on
Conference_Location :
Yunnan
Print_ISBN :
978-1-4244-9712-6
Electronic_ISBN :
978-0-7695-4335-2
DOI :
10.1109/CSO.2011.41
