Generic Lumped Parameter Thermal Model with Optimized Use of Computational Resources

In this paper, a method for efficient computation of generic lumped parameter thermal models is presented. It describes the implementation of thermal model to a given system with a complexity scalable to the available resources. The applied optimization techniques for model computation will be discussed. A performance analysis of the model simulation, regarding the required memory and computing power, will proof that it can be executed in a microcontroller in real-time. It can also be used for overheat prediction under known operating points. A model for an induction machine is presented as an example.