Dynamic simulation for gas distillation column based on an improved mechanism model

A rigorous dynamic mechanism model was developed for the gas distillation column. To improve the rigorous dynamic model's computing efficiency, a novel kernel ridge regression algorithm was adopted to predict the vapor-liquid equilibrium data. In order to solve the differential-algebraic equations, the Gear's method was considered and the computing results based on steady-state mechanical model were used as the initialization value of the rigorous dynamic model. Based on this dynamic model, the distillation process had been simulated and analyzed. The simulation results showed that the dynamic model was consistent with the actual experiment value and the model's computing efficiency was improved. The efficient dynamic model could be applied for developing the operator training simulator of the gas distillation process.