This study introduces a novel robust optimal control design for a flexible joint robot manipulator. The proposed control design combines robust optimal control (ROC) and robust optimal integral sliding mode control (ROISMC), transforming the robust control problem into an optimal control problem for both matched and mismatched unknown functions, as demonstrated through the Lyapunov function. The validity of both schemes is established through comparison with the desired results of flexible joint angle and rotor joint angle and velocity. It is found that their results are accurately alignment with the desired trajectory, even in the presence of mismatched functions.