Generic machine identification and maximum efficiency operation of induction machines

TL;DR

This paper introduces a universal method for identifying induction machine parameters and optimizing their operation for maximum efficiency using steady-state rotor flux measurements, applicable even without torque sensors.

Contribution

It presents a general identification approach that accounts for nonlinear effects and enables maximum efficiency operation in inverter-fed induction machines.

Findings

Achieves maximum efficiency per torque using lookup tables.

Comparable results obtained without torque sensors.

Quantitative comparison of V/Hz and FOC efficiencies.

Abstract

This paper proposes an advanced machine identification method for inverter fed squirrel-cage induction machines, based on steady-state measurements in the rotor flux oriented dq-reference frame. The measured machine maps are used to extract maximum efficiency per torque (MEPT) look-up tables, which guarantee the maximum achievable efficiency in every operating point. Furthermore, it is shown, that comparable results can be achieved, even without a torque sensor. The main advantage of the described method is its generality, which implicitly covers magnetic saturation, iron losses and other nonlinear effects that are typically neglected or approximated by complex models. Finally, the efficiencies of V/Hz and field-oriented control (FOC) are calculated for different speeds and load torques, allowing for quantitative assessment and comparison of both methods.