Methodology for Designing Embedded Real-Time Electrothermal Models in PYNQ Z1 System on Chip

This paper presents a design methodology for Real- Time Digital Twin Electrothermal Models on SoC (Systems on a Chip) developing an analysis of the most relevant constraints for Real-Time Operation and their classification. Moreover, the electrothermal model layout within the architecture of the PYNQ Z1 SoC will be presented exposing the advantages of using an external interface such as Jupyter for Forecasting and the role of parallel structures to accelerate the computation of numerical integration algorithms in Programmable Logic (PL). The proposed electrothermal models will be initially validated with simulation results and later compared with experimental data provided by CMCS (Compact Motor Control System), in which 4 asynchronous engines are supplied by 4 DC/AC IPBB (Inverter Power Building Block).