Estimation of time-dependent parameters of a heat-integrated distillation column system

A rigorous model is required for real-time dynamic optimization and nonlinear model predictive control. The estimation of time-dependent parameters of such a model has been a challenging task. In this work, we propose a method to solve this problem by dynamic optimization. Time-dependent tray efficiencies and the unmeasured feed composition in a heat-integrated distillation column system are estimated based on the measured data of temperatures and flow rates. Adjustable weighting factors are introduced to enhance the identifiability of the multi-parameter system. The results show a satisfactory estimation by using this approach.