Optimal Stopping with Random Maturity Under Nonlinear Expectations