Nonlinear Wave-Based Measurement Location Design and Online Estimation of a High-Purity Heat-Integrated Distillation Column

A heat‐integrated distillation column (HIDiC) provides a great opportunity for energy conservation in distillation. The nonlinear wave theory offers an accurate and simplified description of the HIDiC. To utilize the wave model online, an estimator design is necessary since the wave model parameters cannot be measured directly. A method for selecting the location of the stage concentration measurements is proposed here to estimate the wave model parameters online. The first several measurement locations are selected based on a weighted principal component analysis (PCA) method, and the remaining measurements are decided on by a relative distribution index. Separation of a benzene‐toluene mixture is used to illustrate the performance of the selection and estimation. The simulation results show that the proposed method has a stronger adaptability than the method based on PCA. A weighted principal component analysis method and a relative distribution index are combined to form an approach for selecting the measurement locations in a heat‐integrated distillation column. An online estimator is then designed based on the selection method and the nonlinear wave model. The stage concentrations can also be predicted.