430115 Moving Horizon Based Real Time Optimization and Advanced Hybrid Model Predictive Control of Continuous Pharmaceutical Manufacturing Process

Thursday, November 12, 2015: 5:20 PM
Salon D (Salt Lake Marriott Downtown at City Creek)
Ashish Shah, Rohit Ramachandran and Ravendra Singh, Chemical and Biochemical Engineering, Rutgers University, Piscataway, NJ

The pharmaceutical company is still largely a batch process industry. The strict regularity requirements from the regulatory agencies and different level of complexities involved with pharmaceutical processing has prevented the transition towards continuous manufacturing. However, Quality by Design (QbD), a novel initiative proposed by regulatory agencies, is a big leap in making this transition successful. QbD involves the use of process analytical technology (PAT) tools to monitor the process with a real time feedback system to provide a control strategy suitable for continuous manufacturing. Traditional control strategies such as proportional-integral-derivative (PID) and model predictive control (MPC) have been widely used in chemical and petrochemical industries in the past decades. However, pharmaceutical industries have been slow in adapting this technology in the traditional quality-by-testing (QbT) approach largely due to strict regulatory requirements.

In this work, a moving horizon based real time optimization (MH-RTO) technique has been integrated with a hybrid model predictive control (MPC) system. The integrated system has been applied in a continuous downstream API separation and purification process as well as tablet manufacturing process. The API separation and purification process consist of four continuous downstream unit operations (Crystallization, filtration, drying and mixing) while the continuous direct compaction tablet manufacturing process consist of four unit operations (feeders, co-mill, blender, tablet press). In the proposed approach, the integrated MH-RTO provides the optimal operational set points for the total production rate in real time. The MH-RTO takes into consideration the capital and operating cost, the market fluctuations, the product inventory, the product quality assurance strategy, the regulatory constraints, and the product rejections. An advanced hybrid model predictive control system then ensures that required production rate with desired quality is met with minimum resources and time. A robust optimization strategy and an efficient control system have been integrated to achieve the maximum profit. The MH-RTO integrated with a hybrid control strategy ensures the maximum possible profit irrespective of the market demand fluctuations. The basic advantage of the MH-RTO framework is that it takes into consideration the future demand and thus can lead to increased profit compared to a standard Real Time Optimization approach.

The objective of this presentation is two-fold. First to highlight the moving horizon based real time optimization technique and MPC-PID hybrid control architecture, and second to demonstrate its applications through two case studies: downstream API separation and purification process and tablet manufacturing process.

Extended Abstract: File Not Uploaded
See more of this Session: Economics and Process Control
See more of this Group/Topical: Computing and Systems Technology Division