研究一类可用Ｈａｍｍｅｒｓｔｅｉｎ模型描述的ＭＩＳＯ非线性系统的优化控制方法，提出一种基于前向神经网络的Ｈａｍｍｅｒｓｔｅｉｎ模型，以描述ＭＩＳＯ非线性对象，并给出具有内模结构的优化控制器设计方法，从而实现了在稳态优化与动态优化双重指标下的设定值闭环优化控制。针对汽油管道调合系统进行的仿真实验，验证了该方法的有效性。 A kind of optimal control method of MISO nonlinear system which can be described by Hammerstein model is studied. A Hammerstein model based on feedforward neural network is proposed to describe MISO nonlinear objects. An optimal controller design method with internal model structure is given. Thus, the closed-loop optimal control of setpoint under the dual targets of steady-state optimization and dynamic optimization is realized. The simulation experiment on gasoline pipe blending system verifies the effectiveness of the method.