461728 Surrogate-Based Optimization Methodology for Pharmaceutical Tablet Manufacturing Processes

Thursday, November 17, 2016: 8:30 AM
Monterey I (Hotel Nikko San Francisco)
Zilong Wang, M. Sebastian Escotet-Espinoza, Ravendra Singh, Fernando J. Muzzio and Marianthi G. Ierapetritou, Department of Chemical and Biochemical Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ

The pharmaceutical manufacturing industry is going through a transformation from traditional batch processes to advanced continuous processes with the aim of increasing economic efficiency and process robustness [1]. Such a revolutionary change can further be facilitated by the enhanced process understanding with the help of Process Systems Engineering (PSE) tools [2]. Recently an increasing amount of research has been focused on the modeling of unit operations of continuous pharmaceutical processes, including feeders, blenders, roller compactors, the tablet press, etc. [3] [7], which can then be integrated into a flowsheet model to predict dynamic behaviors of the process under specific operating conditions and material properties. The flowsheet models can then be used in the process optimization problems.

There are several challenges faced by the pharmaceutical manufacturing processes in terms of the process optimization. First, the randomness in nature of the numerous input material properties can make the optimization problem rather complicated. Some inevitable variations in the process (e.g. refill in the feeder) and the propagation of such variations in the subsequent unit operations must be considered in order to insure desired product qualities [4]. Second, due to the strict standards of final pharmaceutical products, it is usually desired to design a rather flexible process that can withstand the potential uncertainty in the inlet materials [5]. Finally, depending on the different modeling techniques that are adopted, the flowsheet model can be computationally expensive and may give noisy output predictions, which makes it difficult to use traditional derivative-based solvers to find the optimal solution [6].

To address the difficulties outlined above, we develop a surrogate-based optimization methodology in order to solve the optimization problems in pharmaceutical manufacturing processes. To prevent considering unnecessary input factors, which have very little impact on the main output variables, a global sensitivity analysis is initially conducted to identify the subset of most significant input factors. Surrogate models (e.g. Kriging, RBF) are built to smooth out the noise in the output predictions, and used as an efficient approximation of the original flowsheet model. The surrogate models are adaptively updated both considering the feasibility of the constraints and the optimality of the objective function, in order to reduce the total sampling cost and efficiently find the optimal solution. The methodology is illustrated using a direct compression continuous manufacturing line.



[1] Plumb K. Continuous processing in the pharmaceutical industry: changing the mind set[J]. Chemical Engineering Research and Design, 2005, 83(6): 730-738.

[2] Gernaey K V, Cervera-Padrell A E, Woodley J M. A perspective on PSE in pharmaceutical process development and innovation[J]. Computers & Chemical Engineering, 2012, 42: 15-29.

[3] Rogers A J, Hashemi A, Ierapetritou M G. Modeling of particulate processes for the continuous manufacture of solid-based pharmaceutical dosage forms[J]. Processes, 2013, 1(2): 67-127.

[4] Boukouvala F, Ierapetritou M G. Surrogate-based optimization of expensive flowsheet modeling for continuous pharmaceutical manufacturing[J]. Journal of Pharmaceutical Innovation, 2013, 8(2): 131-145.

[5] Rogers A, Ierapetritou M. Feasibility and flexibility analysis of black-box processes part 2: Surrogate-based flexibility analysis[J]. Chemical Engineering Science, 2015, 137: 1005-1013.

[6] Sen M, Rogers A, Singh R, et al. Flowsheet optimization of an integrated continuous purification-processing pharmaceutical manufacturing operation[J]. Chemical Engineering Science, 2013, 102: 56-66.

[7] Hsu S H, Reklaitis G V, Venkatasubramanian V. Modeling and control of roller compaction for pharmaceutical manufacturing. Part I: Process dynamics and control framework[J]. Journal of Pharmaceutical Innovation, 2010, 5(1-2): 14-23.

Extended Abstract: File Not Uploaded
See more of this Session: Design and Operations Under Uncertainty I
See more of this Group/Topical: Computing and Systems Technology Division