Design and implementation of neural network adaptive inverse control

Bioprocess has the characteristics of complicated dynamics, nonlinearity and multivariable coupling. The main control goal is to get a pure product with a high concentration, which commonly is difficult to achieve by designing traditional controller. To improve the control performance, neural network adaptive inverse (NNAI) controller is proposed on the basis of neural network inverse (NNI) controller to control the mycelia concentration and the substrate concentration in fermentation process. The reversibility of the multi-variable fermentation system is analyzed and neural network is offline trained to approximate the inverse system. Based on the basis functions and Lyapunov theory, an adaptive law is designed to adjust the parameters of neural network in control process. Simulation results show that the control performance of NNAI controller proposed is superior to NNI controller.