The proposed control scheme is easier to design because the pre-computation of the bound on the NN approximation errors by tril and error is not needed.

• 由于不需要对神经网络的逼近误差进行事先计算，本文提出的控制方案十分易于控制实现。