Accurate Magnetic Characterization Based Model Development for Switched Reluctance Machine

Abstract

High performance control and analysis of switched reluctance machines (SRMs) require accurate modeling of their magnetic characteristics. However, the doubly salient structure and deep magnetic saturation make it very complicated to accurately model SRMs. This paper presents a high fidelity model development for SRMs. The model is developed based on the experimental measurement of flux-linkage and torque characteristics. The introduced measurement noises / errors are investigated carefully. Then several post-processes are achieved to reduce these noises. The measurement accuracy is verified by three methods: finite element method (FEM), search coil comparison, and LCR meter. The measured data are employed after proper rearrangement to build a dynamic MATLAB simulation model for the tested 8/6 machine. The model accuracy and dependability is achieved experimentally.

Numerical Estimation of Switched Reluctance Motor Excitation Parameters Based on a Simplified Structure Average Torque Control Strategy for Electric Vehicles

“…The flux-linkage λ(i,θ), inductance L(i,θ), and torque T(i,θ) are functions of both current magnitude (i) and rotor position (θ). The voltage equation is given as follows [25]:…”

“…For each desired rotor position (θ), a pulsed dc voltage is applied to the phase winding; then, v k and i k are measured and recorded. Equation 10is used to estimate the flux from the measured voltage and the current data [25].…”

“…As noted, both the flux and torque data are highly nonlinear functions of the current and rotor position. These data are employed in the form of look-up tables to represent the machine using MATLAB Simulink [25]. The geometrical dimensions of the studied SRM are gathered in Table 1.…”

“…Moreover, the manufacturing process introduces tolerances that make the exact determination of machine geometry data a difficult task. Even with small tolerances, a significant error between FEA and measured results can occur [25]. In [11], the machine model is developed using a Fourier approximation for the inductance profile that has a questionable level of accuracy.…”

“…It does not require any machine geometry data. Besides, the physical effects and imperfections during the manufacturing processes are contained in the measured data [24,25]. Therefore, the experimental measurement of magnetic characteristics is adopted in this paper.…”