Exact penalty functions with multidimensional penalty parameter and adaptive penalty updates

TL;DR

This paper develops a comprehensive theory for exact penalty functions with multidimensional penalty parameters, enhancing flexibility and performance in infinite-dimensional optimization problems through adaptive penalty updates.

Contribution

It introduces a general framework for vectorial penalty parameters, provides conditions for their exactness, and proposes new algorithms for analyzing and ensuring global exactness.

Findings

Vectorial penalty parameters improve flexibility in constraint handling.

New conditions for local and global exactness are established.

A novel algorithmic approach characterizes global exactness via sequence behavior.

Abstract

We present a general theory of exact penalty functions with vectorial (multidimensional) penalty parameter for optimization problems in infinite dimensional spaces. In comparison with the scalar case, the use of vectorial penalty parameters provides much more flexibility, allows one to adaptively and independently take into account the violation of each constraint during an optimization process, and often leads to a better overall performance of an optimization method using an exact penalty function. We obtain sufficient conditions for the local and global exactness of penalty functions with vectorial penalty parameters and study convergence of global exact penalty methods with several different penalty updating strategies. In particular, we present a new algorithmic approach to an analysis of the global exactness of penalty functions, which contains a novel characterisation of the global exactness property in terms of behaviour of sequences generated by certain optimization methods.