Efficient nonlinear predictive control of a biochemical reactor using neural models

Abstract

This paper describes the application of artificial neural networks to modelling and control of a continuous fermentor. A computationally efficient nonlinear model predictive control (MPC) algorithm with nonlinear prediction and linearisation (MPC-NPL) which needs solving on-line a quadratic programming problem is developed. It is demonstrated that the algorithm results in closed-loop control performance similar to that obtained in nonlinear MPC, which hinges on full on-line non-convex optimisation. The computational complexity of the MPC-NPL algorithm is shown, control accuracy and robustness are also demonstrated in the case of noisy measurements and disturbances affecting the process.

Hybrid intelligent control of substrate feeding for industrial fed-batch chlortetracycline fermentation process

“…Instead, they rely on exploiting a large process database. The datadriven models provide the possibility of developing advanced control methods for fed-batch control, such as ANN based MPC methods [19][20][21]. Although various mechanistic and data-driven models have been developed for bioprocesses, both of them have certain disadvantages.…”

“…Numerous literature works have reported applications of advanced control in fermentation processes [2,3,7,8,[18][19][20][21][22][23], involving dynamic programming, online adaptive control, nonlinear optimization, nonlinear control, optimal control, multivariable control, model reference control, fuzzy control, and model predictive control (MPC), etc. These types of methods have gained increasing popularity because of their strong capability in dealing with process nonlinearity, dynamics and optimization.…”