非锥凸最优化问题中的可行距离

Feasible Distance in Non-conic Convex Optimization Problems

当一般的非锥凸最优化问题或其对偶不可行时,通过引入可行距离这一概念,讨论新系统的可行性,并考察了在新系统中可行距离的性质,得到了与其等价的可计算的优化形式.

When a non-conic convex optimization problem or its dual was infeasible, we considered the feasibility for the new system by introducing a concept of feasible distance. Moreover, we investigated the properties and obtained the equivalent optimization formulations with tractability for the feasible distance.