Methodology for Calculation of Efficiency Maps for Permanent Magnet Synchronous Motors from Sparse Measured Data

Abstract

This paper presents a methodology to evaluate the efficiency map of a permanent magnet synchronous machine (PMSM) from sparse measured data. The methodology is based on a model of losses of the machine. Based on the model the efficiency at an arbitrary point can be calculated and an error between the calculation and measurement can be determined. The mathematical model is adapted in such a way that the error between the model and experiment is minimized. The model has from 7 to 20 adapted parameters. The search for the optimal parameters is done with two approaches: brute force and Monte Carlo search. The results are compared in the paper. With the Monte Carlo search the algorithm finishes in few minutes, based on the number of parameters. The result is an analytical function for the efficiency map and the efficiency can be calculated at an arbitrary point with arbitrary precision. The resulting fit between the measured and calculated map is within a few percent difference.