Modeling and Control of Industrial Fischer–Tropsch Synthesis Slurry Reactor Using Artificial Neural Networks

JOURNAL OF CHEMICAL ENGINEERING OF JAPAN

Jianhong Lu¹  Wenguo Xiang¹  Xiaocen Xue¹ 

[1] Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education, Southeast University, Nanjing, Jiangsu 210096, China

关键词: Fischer–Tropsch Synthesis;    Slurry Reactor;    Radial Basis Function;    PID Neural Network;    Differential Evolution;   

DOI  :  10.1252/jcej.14we095

来源: Maruzen Company Ltd

【 摘 要 】

References(18)This study presents an artificial neural network (ANN) approach for the modeling and control of the Fischer–Tropsch synthesis (FTS) slurry reactor. Operating data collected from an FTS demonstration plant were used to develop a radial basis function neural network (RBFNN) model, which is used for predicting the reactor temperature under industrial operation conditions. Additionally, a modified PID neural network (MPIDNN) control method was proposed for the reactor temperature control based on the trained RBFNN model. The differential evolution (DE) algorithm was used as the learning algorithm to automatically optimize the RBFNN and the PIDNN parameters. In the FTS slurry reactor simulation, the RBFNN model achieved satisfactory predictions of the reactor temperature, whereas the MPIDNN control system demonstrated an impressively stable and rapid control of the reactor temperature.

【 授权许可】

Unknown