Sensorless Backstepping Control of Multiphase Induction Machines Under Fault Conditions

This paper deals with the sensorless backstepping control of the multiphase induction motor under faulty operation mode. Indeed, the proposed control is based on a linearized model where decoupling problems associated with the nonlinear dq model are eliminated. Whereas, this model is obtained from the linearization of the machine vector model in the stationary reference frame $\alpha-\beta$ which is achieved by using a non-linearized transformation such as the Z-transformation proposed in this paper. Furthermore, this model can be extended to all planes $\alpha(\text{n})-\beta(\text{n})$, where the used backstepping technique allows ensuring independent torque and rotor flux stabilization in each plane. Finally, some experimental results are presented for the validation of the proposed control on a specific sensorless benchmark of a five-phase squirrel cage induction motor under fault conditions suchas open phase fault on one and two phases.