Crystallization process control has attracted extensive research interest in recent years within the field of industrial crystallization. This study focused on controlling the crystallization process of coupling cooling and antisolvent crystallization (CCAC), a traditional crystallization method widely used for enhancing the product yield and process efficiency. By employing model-free control, real-time self-feedback control, and preset process routes optimized by genetic algorithm (GA), the simultaneous control of the temperature profile and antisolvent addition was achieved during CCAC. The kinetic behavior of l-carnitine during the CCAC process under different control logics was characterized. It is revealed that the real-time self-feedback control and preset process route based on genetic algorithm exhibited better performance than the model-free control strategy and linear process. Specifically, real-time optimization based on the genetic algorithm (GA-RTO) can achieve a larger D90, while the optimized design route of the process can result in a larger D50. However, for simple process development, a preset process path may yield acceptable preliminary results.

中文翻译：

---

实时优化和遗传算法，用于增强耦合冷却和反溶剂结晶的过程控制


近年来，结晶过程控制在工业结晶领域引起了广泛的研究兴趣。本研究的重点是控制耦合冷却和反溶剂结晶 （CCAC） 的结晶过程，CCAC 是一种广泛用于提高产品产量和工艺效率的传统结晶方法。通过采用无模型控制、实时自反馈控制和遗传算法 （GA） 优化的预设工艺路线，在 CCAC 期间实现了温度曲线和反溶剂添加的同步控制。表征了不同控制逻辑下 l-肉碱在 CCAC 过程中的动力学行为。结果表明，实时自反馈控制和基于遗传算法的预设过程路线表现出优于无模型控制策略和线性过程的性能。具体来说，基于遗传算法 （GA-RTO） 的实时优化可以实现更大的 D90，而工艺的优化设计路线可以产生更大的 D50。但是，对于简单的工艺开发，预设的工艺路径可能会产生可接受的初步结果。