Modeling and adaptive control for batch sterilization

Abstract

This contribution addresses the control problem associated with tracking temperature profiles in batch sterilization processes. Dynamic variability of the plant and the unsteady state heat transfer to the product are shown to cause a significant performance degradation when simple PID-type controllers are employed to control the operation. This is especially true when operating at high temperature and low energy consumption. To improve performance, several adaptive control techniques such as selftuning regulator, stable adaptive control and adaptive IMC are compared on a simulation background, and their advantages and drawbacks discussed. In particular, it is shown that performance degradation often occurs at the early stages of recursive estimation, resulting in product overprocessing. To overcome such control degradation, a priori information derived from conservation principles of mass and energy, is incorporated into the control structure and combined with classical recursive identification techniques. This will avoid the appearance of high parameter uncertainty that otherwise would cause degradation of the closed loop response and/or instability. The complete scheme will be included into the IMC framework and their efficiency demonstrated on tracking sets of constant as well as variable (optimum) set point profiles for batch thermal processing.