Extend constraint dissolving mappings to general set-membership constraints

Resolve the technical challenges caused by the nondifferentiability of the projection operator $P_C(G(x))$ in order to construct and analyze constraint dissolving mappings for general set-membership constraints $G(x)\in C$, rather than restricting the framework to equality constraints with $C:=\{0\}$.

Background

The proposed constraint dissolving framework is formulated for optimization problems with general set-membership constraints G(x)CG(x)\in C, where CC is a closed convex set, and geometric constraints xDx\in D, where DD may be nonconvex. However, the explicit construction of the constraint dissolving mapping relies on the projection PC(G(x))P_C(G(x)).

Because this projection can be nondifferentiable in the general set-membership setting, the paper’s concrete construction and numerical analysis are restricted to equality constraints, corresponding to C:={0}C:=\{0\}. Extending the construction to general closed convex sets would broaden the applicability of the method beyond the equality-constrained cases treated in the implementation section.

References

However, the nondifferentiability of $P_C(G(x))$ poses the significant technical challenges that remain unsolved currently.

A constraint dissolving inexact penalty method for optimization problems with geometric constraints  (2608.24542 - Jia et al., 25 Aug 2026) in Section 4, Implementation of constraint dissolving mappings for nonconvex sets