A control method of substrate feeding about lysine fermentation

Abstract

A fuzzy neural network inverse model is established in order to solve the optimal and the maximum output rate in lysine substrate feeding fermentation process control through research the structural and parameters. The model is more robust, more adjusts the membership function automatically and more dynamics in the rule optimal control than the traditional rule-based fuzzy control. And it is trained by the optimal production data in the actual process of lysine substrate feeding. The output of the substrate feeding inverse model is the real-time input of system. Experimental results show that lysine productivity improved significantly and achieve real-time online control by the method in lysine substrate feeding process control.