相对扰动增益矩阵是分析控制结构抗干扰性的一个有用工具,但它的结果受模型不确定性影响。提出基于优化来确定相对扰动增益矩阵范围的方法并用来选择ALSTOM气化装置的控制结构。利用过程运行仿真数据和输出误差系统辨识方法得到降阶线性模型。在不同运行状况统辨几个模型并通过这些模型得到模型不确定性范围。计算出ALSTOM气化装置在模型不确定性下的广义相对扰动增益范围并用来选择鲁棒控制结构。 Relative disturbance gain matrix is ​​a useful tool to analyze the anti-interference of control structure, but its result is affected by model uncertainty. A method to determine the relative disturbance gain matrix range based on optimization is proposed and used to select the control structure of the ALSTOM gasification unit. The reduced-order linear model is obtained by using the simulation data of process operation and the system identification method of output error. Several models are discerned in different operating conditions and the model uncertainty range is obtained from these models. The generalized relative perturbation gain range of the ALSTOM gasification unit under model uncertainty is calculated and used to select a robust control structure.