Spray Fluidized-Bed Granulation Method Phase for Drug Development: Control of Particle Quality via Data-Driven, Model-Free Adaptation


A novel data-driven model-free adaptive control (DDMFAC) approach is first presented, and then its stability and convergence analysis is provided to show algorithm stability and asymptotical convergence of tracking error. This approach combines the advantages of model-free adaptive control (MFAC) and data-driven optimal iterative learning control (DDOILC). In order to determine the occupied proportions of MFAC and DDOILC in accordance with their diverse control performances in various control stages, fuzzy logic is also used to adaptively adjust the parameters of the suggested approach. The proposed fuzzy DDMFAC (FDDMFAC) approach is utilised to regulate particle quality in the spray fluidizedbed granulation process's drug formulation stage. This approach's control efficiency is contrasted with that of MFAC, DDOILC, and their fuzzy versions, in which MFAC and DDOILC's parameters are adaptively changed using fuzzy logic. The efficiency of the suggested FDDMFAC approach is verified by a number of simulations.