【摘要】 针对微生物发酵过程中菌体细胞浓度难以实时在线检测的问题,提出了一种基于径向基函数(RBF)神经网络的软测量模型,采用了可调基函数宽度的计算方法,提高了RBF网络的自适应性及泛化能力,为复杂系统中生物量参数的检测提供了一条有效途径。仿真结果表明了该方法的可行性和有效性。更多还原

【Abstract】 A soft-sensing model based on RBF neural network has been proposed against the on-line detection of cell concentration in microbe-fermentation process , which is difficult to accomplish. A computing method of adjusting the basis function width has been applied in order to improve the self-adaption and generalization of RBF network. It can solve the problem of on-line biomass parameter estimation in the complicated system. The feasiblity and reliability of this method is shown by simulation experiments.更多还原