催化裂化过程建模与稳态优化控制

催化裂化装置是一个高度非线性、时变和长时延、强耦合、分布参数和不确定性的复杂系统.在研究其过程机理的基础上,定义了一种模糊神经网络用以建模,用自相关函数检验法检验模型的正确性,再用改进的Frank-Wolfe算法进行稳态优化计算.并以一炼油厂催化裂化装置为对象进行试验,研究其辨识、建模和稳态优化控制.

Modeling and Stable State Optimal Control for FCC Process

Fluid catalysis and cracking unit (FCCU) is a complex system with highly non-linear, time variable, long time delay, intensive coupling, parameter distributed and indefinite. According to the research on the process mechanism of the system, a fuzzy neural network (FNN) was established for the modeling. The correctness of the model was tested with the autocorrelation function checking method, the stable state optimization was computed with improved on Frank-Wolfe algorithm. Taken an example for the FCCU of a oil refinery, the system identification, modeling and stable state optimal control was studied and tested.