Robust time-sub-optimal control of the saturated double integrator applied to attitude stabilization