Search

Field-oriented control (FOC) is standardly used in high-performance induction machine (IM) applications. However, the performance of FOC algorithms strongly depends on the level of agreement between the actual IM and the respective model used for the algorithm design. IM parameters’ values are known to vary with the operating conditions, so their online identification is preferable. In addition, it is often preferred to acquire the rotor speed by estimation because speed sensors increase the cost and size of the system while reducing its reliability. Model-reference-adaptive-system (MRAS) estimators are popular due to their simple implementation, but they are sensitive to IM parameters and, hence, usually include a parameter tuning mechanism. This article proposes an indirect rotor-field-oriented controller with the parallel MRAS speed estimation and IM parameter tuning – implemented in an induction generator system. Both the control algorithm and the MRAS include the stray load and iron losses, magnetic saturation, and rotor resistance variation, whereas the MRAS additionally accounts for variations of the stator and stray-load resistances. The system performance is experimentally evaluated over wide ranges of the IM flux, speed, torque, and temperature. [ABSTRACT FROM AUTHOR]

Standard vector controllers for induction machines (IMs) are based on the conventional IM model. Consequently, they are unaware of the existence of stray load and iron losses. However, these losses may induce a considerable mismatch between the actual and reference flux vectors in vector controllers. This mismatch is greatly affected by the operating flux and load. In this paper, two different approaches are considered for compensation of this mismatch—the first involves investment in IMs with low stray load and iron losses, whereas the other involves development of a control algorithm in which both these losses are compensated. The performance of the developed vector controller is compared to the standard controller and tested on three small IMs of different efficiency class and rotor cage material (aluminum versus copper). Simulation and experimental testing is carried over wide ranges of loads, speeds, and magnetization levels. The estimated rotor speed is used as a measure of the controller detuning. A standard estimator that utilizes rotor flux-linkage estimates within the model reference adaptive system is here for the first time modified to include stray load and iron losses. Finally, problems regarding numerical integration, sampling frequency, and temperature-induced variations of IM parameters are addressed. [ABSTRACT FROM AUTHOR]