A Simple Least Squares Approach for Low Speed Performance Analysis of Indirect FOC Induction Motor Drive Using Low-Resolution Position Sensor

Abstract

This paper investigates an ordinary least squares based indirect field oriented control (IFOC) of induction motor at very low speed range (1-5 rpm) using a low-count encoder. At very low speeds the speed estimation provided by common techniques for low-count encoder applications such as averaging the last two speed samples and/or considering the previous speed value until the next position update can generate erroneous speed results which can cause instability in the overall control system. By proposing ordinary least squares (OLS) method with low-resolution position sensor based speed estimation for IFOC of IM drives more realistic speed can be obtained by fitting the predefined sets of speed data in a linear or quadratic manner. Experimental tests prove that the proposed scheme which is constituted of OLS (linear) speed estimation algorithm increases stability and improves low and oscillating speed performance of IFOC of IM drive with low-count encoders down to four pulses per revolution.