A CLASS OF STOCHASTIC SECOND-ORDER-CONE COMPLEMENTARITY PROBLEMS

摘要

The second-order-cone programs constitute an important class of convex optimization problems and as their Karush-Kuhn-Tucker conditions, the second-order-cone complementarity problems have attracted much attention of researchers. In this paper, we consider a class of stochastic second-order cone complementarity problems (SSOCCP). Since uncertainty often occurs in practice, we can find many applications of the considered SSOCCP. With the help of some complementarity functions, we transform the SSOCCP as a stochastic programming problem and by employing some approximation techniques, we present an approximation method. Under some moderate conditions, we establish a comprehensive convergence analysis. Furthermore, we derive some results related to exponential convergence rate and error bounds. Finally, we report some numerical experiments on a chance-constrained optimal power flow problem.