(185g) Adaptive Nonlinear Model Predictive Control of a Continuous Direct Compaction Tablet Manufacturing Process

Active control strategies are playing a vital role in modern pharmaceutical manufacturing. Indeed, digitalization and automation are revolutionizing the pharmaceutical industry and are accelerating the shift from batch operations to continuous operations. To deal with variations in raw material properties and process uncertainties as well as to guarantee desired critical quality attributes (CQAs) of in-process materials and final products, active control strategies are needed to provide real-time corrective actions when departures from quality targets are detected. Instead of applying traditional control strategies by controlling process input variables at fixed setpoints or within tight ranges, active control strategies serve as a “must-have” technology, as mentioned in the recent AIChE PD2M Advanced Process Control workshop.¹

A three-level hierarchical control structure has been applied in the pilot plant at Purdue University in order to achieve effective setpoint tracking and disturbance rejection in the tablet manufacturing process.² Level 0 control is equipment built-in control, Level 1 control is process analytical technology (PAT)-based PID control, and Level 2 control is optimization-based model predictive control (MPC). As can be expected, when the multiple-input multiple-output (MIMO) system exhibits strong process variables interaction, MPC demonstrates better control performance than classical PID control. That is because MPC can predict the transition of controlled variables (e.g. API content, tablet weight) instantaneously and optimize the control moves of all the manipulated variables (e.g. API flow rate, depth of filling) within a finite time horizon. Several MPC implementation case studies have been reported: (1) the API concentration can be controlled in the feeding-blending unit,³ (2) the compaction force and the tablet weight are controlled in a rotary tablet press, ^4-5 and (3) the hopper level of the tablet press and the API concentration are controlled in a direct compaction line.⁶ While satisfactory control performance is found in the aforementioned research studies, the implementation of MPC in actual pharmaceutical manufacturing processes is still in the infant stage.

This work is focused on developing and implementing nonlinear model predictive control (NMPC) in continuous tablet manufacturing. To mitigate the control degradation caused by plant-model mismatch, careful monitoring and continuous improvement strategies are required. When moving horizon estimation (MHE) is integrated with NMPC, historical data in the past time window together with real-time data from the sensor network, enable state estimation and model parameter updating. The adaptive model in the NMPC strategy can compensate for the process uncertainties, further reducing plant-model mismatch effects. The nonlinear first-principles model used in both HME and NMPC can predict the essential but complex powder properties and provide physical interpretation of abnormal events. In the continuous direct compaction process studied in this work, CSTR-in-series models are used to describe the dynamics of the feeding-blending unit and the Kawakita model accounts for powder compressibility in the tablet press. The adaptive NMPC implementation and its real-time control performance analysis will be demonstrated on our continuous tableting pilot plant.

References

[1] Huang, J., O'Connor, T., Ahmed, K., Chatterjee, S., Garvin, C., Ghosh, K., ... & Warman, M. (2021). AIChE PD2M Advanced Process Control workshop‐moving APC forward in the pharmaceutical industry. Journal of Advanced Manufacturing and Processing, 3(1), e10071.

[2] Su, Q., Ganesh, S., Moreno, M., Bommireddy, Y., Gonzalez, M., Reklaitis, G. V., & Nagy, Z. K. (2019). A perspective on Quality-by-Control (QbC) in pharmaceutical continuous manufacturing. Computers & Chemical Engineering, 125, 216-231.

[3] Singh, R., Sahay, A., Karry, K. M., Muzzio, F., Ierapetritou, M., & Ramachandran, R. (2014). Implementation of an advanced hybrid MPC–PID control system using PAT tools into a direct compaction continuous pharmaceutical tablet manufacturing pilot plant. International journal of pharmaceutics, 473(1-2), 38-54.

[4] Su, Q., Bommireddy, Y., Shah, Y., Ganesh, S., Moreno, M., Liu, J., ... & Nagy, Z. K. (2019). Data reconciliation in the Quality-by-Design (QbD) implementation of pharmaceutical continuous tablet manufacturing. International journal of pharmaceutics, 563, 259-272.

[5] Bhaskar, A., Barros, F. N., & Singh, R. (2017). Development and implementation of an advanced model predictive control system into continuous pharmaceutical tablet compaction process. International journal of pharmaceutics, 534(1-2), 159-178.

[6] Kirchengast, M., Celikovic, S., Rehrl, J., Sacher, S., Kruisz, J., Khinast, J., & Horn, M. (2019). Ensuring tablet quality via model-based control of a continuous direct compaction process. International journal of pharmaceutics, 567, 118457.