Comprehensive temperature characterization of a thermal system using distributed RC networks

The aim of this study was to develop a method for determining the thermal behavior of any thermal system by identifying thermal parameters for multidisciplinary behavioral models. The proposed solution refines the minimum structure of the thermoelectric model using an iterative least squares method and the Levenberg-Marquardt algorithm, and a graph of the spectral density of time constants. This approach reduces the influence of the filtering factor during the deconvolution operation, enabling the construction of a thermal model using the structural function of the thermal characteristic transition. These findings have potential implications for the development of integrated circuit behavioral models that account for thermal processes.