Abstract: The control of distillation column is commonly used in industrial equipment in the petrochemical industry, if there is a big tower parameter disturbance, the control mode of the traditional PID is not ideal in the control effect, but the application effect of the complete form of the multivariable decoupling model is not ideal, because the matching of the control model is not ideal, there is a problem that the parameters do not match the actual application. In this regard, the state space multivariable predictive control method is introduced, and the traditional PID control strategy is combined with the design of the distillation tower controller. Because of the combination of the advantages of the PID controller and the state space multivariable control method, the algorithm studied can effectively improve the control precision and response speed of the distillation tower, and effectively suppress the oscillation problem during the distillation operation. The experimental analysis shows that the proposed algorithm has a better control effect in the control process of the distillation column, and the effectiveness of the algorithm is verified.

Key words: State space, multivariable, error correction, distillation tower, predictive control

摘要： 精馏塔控制是石油化工领域常用的工业设备，如果精馏塔相关参数存在较大的扰动时，PID等传统的控制方式在控制效果上不够理想，但是完全的多变量形式的解耦模型应用效果也不够理想，因为控制模型的匹配度不理想，存在参数与应用实际不相符的问题。对此，引入状态空间多变量预测控制方式，结合传统PID控制策略进行精馏塔控制器的设计，所研究的算法因为融合了PID控制器和状态空间多变量控制方式的优点，可有效提升精馏塔的控制精度和响应速度，有效抑制精馏塔运行过程中的振荡问题。通过实验分析，显示所提算法在精馏塔控制过程中具有更佳的控制效果，验证了算法的有效性。

关键词: 状态空间, 多变量, 误差校正, 精馏塔, 预测控制