An intelligent control system for protein purification based on AI reinforcement learning
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SUZHOU JIUTONG HENGTAI MEDICAL TECHNOLOGY CO LTD
- Filing Date
- 2026-04-17
- Publication Date
- 2026-07-17
AI Technical Summary
Existing protein purification chromatography processes lack the ability to dynamically respond to batch-to-batch differences in biological feed and the degradation of chromatography column performance, resulting in poor consistency in yield and purity. Furthermore, the lack of proactive safety defense mechanisms poses risks of equipment damage and material spoilage.
An AI-based reinforcement learning-based intelligent control system is adopted. Through a state space construction module, a reinforcement learning agent module, a safety projection layer module, and an online training module, multi-dimensional state parameters are collected in real time for dynamic strategy optimization and safety verification. Combined with a lightweight state prediction model and a reward calculation module, adaptive adjustment of process parameters and safety constraints are achieved.
It improves the consistency of target protein yield and purity between different batches, avoids the risk of process overruns caused by strategy trial and error, and ensures system safety under extreme abnormal conditions, preventing material scrap and equipment damage.
Smart Images
Figure CN122410985A_ABST