Optimization of Batch Cooling Crystallization of Sodium Phosphite Through Genetic Algorithm

In this paper, the seeded batch cooling crystallization of sodium phosphite (SP) is simulated and optimized through a coupled method of the genetic algorithm and nonlinear programming. At first, the modeling and simulation test methods of the crystallization process are applied for the crystallization of SP, which expands the relevant study of SP from the experiment to the simulation. A comprehensive model is established in MATLAB/Simulink, and based on this model, the results of the common cooling strategy (linear cooling) on the process are investigated. Meanwhile, the process sensitivity to the change of seeding conditions is analyzed. Then, the coupled optimization method based on the genetic algorithm and nonlinear programming is applied to optimize the crystallization process for the first time, and the obtained optimized cooling strategy is compared to the result of the traditional nonlinear programming method (NLPM). The traditional NLPM has more significant effects on large seeding mass and small mean size, while the coupled method has better adaptability. When the coefficient of variation is almost fixed, the cooling strategy obtained by the coupled method could produce more crystals with large mean size. In addition, the end of the process can be reached earlier. The results show that the coupled method is more suitable for the optimization of the batch cooling crystallization of SP.