Hessian approximation of the Lagrange function for nonlinear optimization algorithm Sℓ1QP method.
山影, 裕紀YAMAKAGE, Yuki
562015-03-24 , 法政大学大学院理工学・工学研究科
This paper describes methods for the Sequential Quadratic Programing(SQP) method and Sℓ1QP. SQP method is known as most effective method solution for solving the constrained nonlinear optimization problem.On the other hand, Sℓ1QP method was suggested by Fletcher. It is the method that adopted trust-region method in SQP method. It can be expected by Sℓ1QP method that the performance of SQP improve more.In this paper we compare SQP method with Sℓ1QP method which implemented BFGS method or SR1 method in numerical results.