INTERNAL MODEL CONTROL FRAMEWORK USING NEURAL NETWORKS FOR THE MODELING AND CONTROL OF A BIOREACTOR

An internal model control (IMC) framework using neural networks for th e modeling and control of a nonlinear bioreactor is presented. Unlike existing IMC design techniques, this approach needs no mathematical mo dels. It is shown that, to obtain an accurate inverse model, one needs to use steady-state data in addition to the transient data for traini ng the networks. An integration of neural networks and an unstructured math model of the bioreactor is also proposed to improve the neural n etworks' modeling accuracy. This hybrid approach shows significantly b etter performance than the ''black box'' method, and almost as good a performance as a nonlinear IMC based on an exact mathematical model. T he hybrid method also has other advantages, such as the use of only st eady-state data and the need for only one neural network that can be u sed for both the process model and the inverse process model. Simulati on results show that the neural-net strategy is superior to a PI contr oller.