Adjoint-based predictor-corrector sequential convex programming for parametric nonlinear optimization. (English) Zbl 1273.49040
This paper proposes and analyses an algorithm to solve parametric nonlinear optimisation problems, where the target function is assumed to be convex and the parameter appear in a linear way in the equality constraints. The proposed algorithm is based on a predictor-corrector path-following approach in combination with sequential convex programming to compute solutions of the parametric nonlinear optimisation problems for a sequence of parameter values. The convergence of the proposed method is analysed for exact as well as for inexact derivative information. The optimal control of a hydro power plant serves as numerical example to illustrate the performance of the new approach.
Reviewer: Andrea Walther (Paderborn)
MSC:
49M37 | Numerical methods based on nonlinear programming |
65K05 | Numerical mathematical programming methods |
90C25 | Convex programming |
90C30 | Nonlinear programming |
90C31 | Sensitivity, stability, parametric optimization |