Bioreactor profile design and optimization for ethanol production

Ethanol is an alternative energy source being used now a daysto replace diesel and petrol. By fermentation of glucose using Saccharomyces yeasts under a controlled environment, ethanol is produced. The temperature, dissolved oxygen (DO) and pH in bioreactor must follow desired profiles to give optimum yield. In this paper the process model for yeast fermentation has been described in terms of temperature, pH and dissolved oxygen. The temperature in bioreactor has been controlled by using nonlinear Autoregressive - moving average (NARMA) neuro controller. The two degree of freedom (2DOF) PID controllers have been used for pH and dissolved oxygen control. Levenberg-Marquardt training algorithm has been used to train NARMA neuro controller. The 2DOF-PID controllers have been tuned by MATLAB´s auto tune feature along with manual tuning. Random profile variations have been used for simulation to study the dependence of ethanol on process parameters. An overall increase was noted in the production of ethanol. The performance of the controllers for following desired random profiles has also been shown in contrast with prior researches. The performance of the controllers for following the desired random profiles has shown overall improvement in controlling the process.