SA-BFO and RBF neural network based soft measurement model for CO2 content purified by methanol

Xiaoyi WANG,Lingbo ZHANG,Xingsheng GU

DOI: https://doi.org/10.16652/j.issn.1004-373x.2016.13.023

2016-01-01

Abstract:In order to solve the problems of the bacterial foraging optimization(BFO)algorithm easily falls into the local optimum,and the reverse direction during chemotaxis operation is uncertain,the characteristic of simulated annealing(SA)al?gorithm is used to propose the simulated annealing and bacterial foraging optimization (SA?BFO) algorithm. SA algorithm can reach the global optimal solution to the maximum extent while obtaining the local optimal solution. The improved algorithm is ap?plied to optimizing the RBF neural network,and establishment the soft measurement model for CO2 content purified by metha?nol. The simulation results show that the model has high accuracy and precision,and has a certain contribution value for greatly improving the methanol production.