Abstract:

To achieve the desired crystal size distribution (CSD) for the cooling crystallization process of β form L-glutamic acid (LGA), the population balance equation (PBE) of size-dependent growth rate was established by using the method of characteristics (MOCH), and then the optimal supersaturation and temperature curves during the crystallization process were determined by parameter estimation for the established PBM. Due to the nonlinearity and non-convexity of the objective function for model parameter identification, a small number of economic batch crystallization experiments were conducted in combination with image analysis on the crystal seeds and product shapes, in order to fit the model parameters. According to the practically required cycling time for crystallization, to realize the target CSD, the optimal curve of decreasing temperature was determined by optimizing the supersaturation curve of the crystallization process, so as to implement the control of crystal growth process based on the maintained supersaturation level. Experiment results showed that the desired CSD could be achieved by implementing the optimal supersaturation control scheme via the optimized temperature control curve.

Key words: population balance equation, crystal size distribution, crystallization, image processing, supersaturation control

摘要：

针对β型L-谷氨酸冷却结晶过程，为获取期望粒度分布，采用特征曲线法（MOCH）来建立关于粒度相关生长率的种群平衡方程（PBE），然后通过对种群平衡模型（PBM）参数辨识后确定最优过饱和度及控温曲线。由于辨识模型参数的目标函数具有非线性和非凸型性，因而采用少量经济性的批量冷却结晶实验，结合图像分析晶种和产品粒度分布得到的统计数据，拟合模型参数。根据实际要求的结晶过程时间，为达到目标粒度分布，通过优化结晶过程的过饱和度获得最优调温曲线，实现基于恒定过饱和度的晶体生长过程优化控制。实验结果表明通过优化的控温曲线，实现了基于最优过饱和度控制的期望目标粒度分布。

关键词: 种群平衡公式, 粒度分布, 结晶, 图像处理, 过饱和度控制