Predicting of Sludge Acidogenic Fermentation Process on the Mathematical Models

The prediction and control for sludge acidogenic fermentation process has a guiding role for the biological treatment of acid mine drainage with SRB to select proper carbon source. Two mathematical models were used to predict sludge acidogenic fermentation process with semi-continuous operation mode. The results showed that quadratic polynomial regression model is superior to multivariate linear regression models. It can predict acetic acid yield rate and total acid productivity well, and the average relative error between predicted value and actual value is less than 10%. But it cannot correctly predict prop ionic and butyric acids yield, and the average relative error between predicted value and actual value is less than 17%, respectively.