Quadrature Axis Flux Modeling in Projected Pole Synchronous Machines Using Regression Techniques and Adaptive Network Fuzzy Inference System
This paper presents new generalized models for estimating quadrature axis (q-axis) magnetic flux in salient pole synchronous machines. The models are designed such that knowing the direct axis (d-axis) flux and the field excitation, q-axis flux of the synchronous machines is estimated. Available experimental flux values on a laboratory machine in d-axis and q-axis measured using Germanium diodes as Flux sensors and the calculated flux values of three different synchronous machines have been used in the present work for the development of an Adaptive Network Fuzzy Inference System (ANFIS) based model and two mathematical models employing Multi-Variable Linear Regression (MVLR) and Multi-Variable Polynomial Regression (MVPR). The models are used to calculate the q-axis flux during field excitation change.