Model Predictive Control Approach to Improve the Control Performance in Pasteurized Milk Process

In a pasteurized milk process, the milk must be heated to above 72C and subsequently cooled to a low temperature about 4C in order to ensure a food safety. According to the real pasteurized milk plant, individually single variable control approach including manual control with butterfly valve and automatic control by on-off controller have been carried out to control milk level at storage tanks, hot water temperature at the boiler, and iced water temperature at the ripple plate, respectively. Although the on-off controllers adequately control the milk temperature at holding tube and cooling section of plate pasteurizer (PP) at process requirements, the dynamic profiles of milk temperature fluctuate around desired value. This leads to insufficient energy consumption for pasteurized milk production. To handle this, a multi-variables control approach with model predictive control (MPC) has been proposed owing to an interaction between manipulated variables. The control performance is simulated based on MATLAB software. The simulation results show that the MPC can control the milk and water temperature at the plant’s requirements with less drastic control action. This shows that the MPC is applicable to control the milk and water temperature. Keyword: Model predictive control, Multi-variables control, Mathematical model, Pasteurized milk process, Energy cost saving TIChE International Conference 2011 November 10 – 11, 2011 at Hatyai, Songkhla THAILAND pc001-2