Estimating parameters of permanent magnet synchronous motor via finite element methods

This paper aims to identify the parameters of permanent magnet synchronous motors under Gaussian white noise disturbances. The mathematical model of permanent magnet synchronous motor is a system of nonlinear coupled stochastic differential equations, which takes d- and q-axes voltages as inputs. The three-phase voltage is amplified with Gaussian white noises to simulate the real environment. The noisy three-phase voltage is then converted into noisy voltages on the d- and q-axes. By setting the d-axis current to 0, we convert the original system into a linear one. The Euler-Maruyama method will be used to obtain the measurement data from the numerical solution of the system. The least square estimation based on the finite element method is proposed to identify the parameters. An estimate of the solution to the system is determined based on some basis functions via the finite element method. Using the least square method, the inductance, flux linkage, stator resistance, rotational inertia and friction coefficient are identified. Experimental results demonstrate that this method can provide more accurate results with fewer measurement data.