Investigation into the adaptive control of a fermentation process

Publisher Website
Google Scholar
Abstract
The availability of a batch-fed yeast fermentation plant provided an opportunity to investigate the performance of an adaptive control scheme for the plant. From observation and measurement of plant outputs, the structure of a model describing the plant dynamics was established. A solution of the model allowed comparisons to be made with the actual plant operation. While the simulation did not reflect the behaviour of the plant precisely it did provide sufficient information to structure a parameter estimation procedure within a state-variable framework. Using the estimator, the multivariable pseudo linear descriptions of the nonlinear model were identified from input and output data and these were passed to a suboptimal controller whose feedback gains were adjusted to minimise a performance index. Results are presented identifying the appropriate model structure and the operation of the multivariable adaptive controller. The work suggests that the application of the adaptive technique to an online configuration will improve the present semiautomated control approach to this problem.

This publication has 1 reference indexed in Scilit: