Monte Carlo and quasi-Monte Carlo sampling methods for a class of stochastic mathematical programs with equilibrium constraints (Q929336)

Considered is the stochastic mathematical program with equilibrium constraints (SMPEC) \[ \min \mathbb E\left[ f(x,y,\omega) \right]\;\text{ s.t. }g(x,y) \leq 0,\;h(x,y) =0; \quad o\leq y\perp \mathbb E\left[ F(x,y,\omega) \right] \geq 0, \] where \(\mathbb E\) denotes the expectation with respect to the random variable \(\omega \in \Omega\), the functions \(f:\mathcal{R}^{n+m}\times \Omega \to \mathcal{R},\;g:\mathcal{R}^{n+m}\to \mathcal{R} ^{s_{1}},\;h:\mathcal{R}^{n+m}\to \mathcal{R}^{s_{2}}\), and \(F: \mathcal{R}^{n+m}\times \Omega \to \mathcal{R}^{m}\) are all twice continuously differentiable, and the symbol \(\perp \) stands for orthogonality of the two vectors on both sides. It is assumed that \(\Omega \) has an infinite number of samples. The authors consider a Monte Carlo sampling method combined with a penalty technique for solving the problem (SMPEC). They establish convergence of global optimal solutions and stationary points of approximation problems generated by the proposed method. Also, they give an example in economics to illustrate the model and show some numerical results with this example.