Non-Linear Optimization-Based Temperature Estimation of IPMSM

In this paper, a non-linear optimization algorithm-based temperature estimation method is presented for interior permanent magnets synchronous motor (IPMSM). Commonly the online temperature estimation scheme is realized by an abstract model with lumped-parameter thermal network. A contribution of this paper is to simplify the model by combining the thermal and loss models with essential parameters. The motor loss calibration can be avoided by this unified model. The parameters are identified by quadratic programming method. Consequently, the identified system can estimate the node temperatures online. This paper presents experimental results based on a forced-coolant IPMSM to validate the proposed method.