A robust data based flux estimator using Neural Network for sensor-less vector controlled IM drives

This paper proposes a robust data based flux estimator for sensor-less vector controlled Induction Motor (IM) drives. The performance of sensor-less vector controlled IM drives to a large extent depends on the accuracy of estimated speed. The speed estimation in turn depends on the accuracy of estimated flux. Conventional voltage model equations for flux estimation encounter major drawbacks at low frequencies/speed. They require a separate estimator to track variations in parameter. This paper proposes a robust flux estimator, designed using single neuron cascaded Neural Network (SNC-NN) to address the major drawbacks of conventional voltage model based flux estimation. The proposed NN based flux estimator is analyzed extensively under various operating conditions, parameter variations and compared with existing voltage model based flux estimator. The promising results obtained are presented.