Estimation of microwave-assisted drying parameters using adaptive optimization inverse techniques

In this work, a parametric adaptive optimization architecture is applied for modelling the direct problem of microwave-assisted drying processes. The proposed architecture, based upon the Levenberg–Marquardt (LM) algorithm, gives a solution for the inverse problem in a complex mathematical model. Experimental results with leather samples in an pre-industrial installation have verified the capabilities of the proposed model for the estimation of optimum non-variable and time-dependent parameters, which are difficult to measure in a real drying process, such as internal evaporation, electric field, and heating-up period. In order to verify the reliability of proposed architecture, some well-known parameters, such as specific heat coefficients, are estimated and contrasted. For parameter and functions estimation, temperature at the centre of the material and moisture content are measured and simultaneously considered. Finally, from analysis of obtained results, some conclusions are extracted with reference to improve the knowledge of internal characterization of microwave-assisted drying processes.