Modeling and parameter estimation for MIP-CGP riser reactor of FCCU

摘要

The modeling of fluid catalytic cracking unit (FCCU) is important due to the significant role of FCCU in oil refining. On account of the strong nonlinearity of FCCU mechanisms the solved optimized parameters are usually local optimum, which is sensible to the parameter initial value. A parameter estimation framework and algorithm is proposed in this paper for getting a stable optimal solution. According to the mechanism network the original model is conversed into several dimension-reduced models to achieve different model parameters respectively. Afterwards, in order to avoid the local optimum, a modified LM algorithm based on a random generated damping factor is proposed to solve the optimization problem resulting from the dimension-reduced models. After some of the initial values of the optimized parameters are determined, the optimization problem can be solved with a simple LM method. The proposed method is applied on the modeling of an industrial FCCU in a petrochemical refinery Co., Ltd. in Jiangxi, China with real industrial operational data. It could get a stable parameters with random initial values within a relative large range. The accuracy property of the resulting model is also verified.