Modeling for liquid-level control system in beer fermentation process

In most beer companies, the productive processes are controlled by means of manul control or open-loop control. The technical equipments are outmoded, the production automation is low and the quality of the products is unstable. Modeling for process objects is very important premise and foundation in order to design fine control system. PLC controller is adopted in the control system. Taking the proportional valve output as the system input, the liquid-level value as the system output, modeling for the two-tank system in beer fermentation process is studied. Based on the sample input and output data by inputting the step signal and pseudo random signal to the system, the deterministic models and stochastic model are identified. The first-order plus deadtime model is obtained using batch least square method in the deterministic model. The ARX model and ARMAX model are chose for stochastic model. Model orders are identified according to Akaike Information Criterion or loss function and other related performance indexes, model parameters are obtained according to the least square method. The two different models are compared and analyzed. An extra experiment is performed to obtain testing data and the comparison result between the model predictive output and the actual output shows the validity of the modeling in the liquid-level control system.