论文部分内容阅读
模型预测控制的经济性能主要通过减少关键工艺参数的方差,以及移动过程的操作点使其更接近约束边界来实现。另一方面,软约束边界需要经常的调整以有效地解决模型预测控制的优化不可行问题。在协调软约束调整与模型预测控制的经济性能的过程中,本文提出了将基于性能评估的最小方差引入到模型预测控制的稳态目标计算中,以保证模型预测控制能够更加合理地提高系统的经济性能。最后,以延迟焦化装置加热炉预测控制为例,讨论和分析了该方法的有效性。
The economic performance of model predictive control is mainly achieved by reducing the variance of the key process parameters and moving the operating point of the process closer to the constraint boundary. On the other hand, the soft constraint boundaries need to be adjusted regularly to effectively solve the problem of optimization of model predictive control. In the process of coordinating the economic performance of soft constraint adjustment and model predictive control, this paper proposes to introduce the minimum variance based on performance evaluation into the steady state target calculation of model predictive control so as to ensure that model predictive control can more effectively improve the system’s Economic performance. Finally, the effectiveness of this method is discussed and analyzed in the case of a delayed furnace predictive control of coking unit.