On the Application of TLS Techniques to AC Electrical Drives

This paper deals with the application of a new neuron, the TLS EXIN neuron, to AC induction motor drives. In particular, it addresses two important subjects of AC induction motor drives: the on-line estimation of the electrical parameters of the machine and the speed estimation in sensorless drives. On this basis, this work summarizes the parameter estimation and sensorless techniques already developed by the authors over these last few years, all based on the TLS EXIN. With regard to sensorless, two techniques are proposed: one based on the MRAS and the other based on the full-order Luenberger observer. The work show some of the most significant results obtained by the authors in these fields and stresses the important potentiality of this new neural technique in AC induction machine drives.