Single Neuron Cascaded Neural Network Model based Speed Estimation for Sensorless Induction Motor Drives

The performance of sensor-less controlled induction motor drive depends on the accuracy of the estimated speed. Conventional estimation techniques being mathematically complex require more execution time resulting in poor dynamic response. The nonlinear mapping capability and powerful learning algorithms of neural network provides an alternative for on-line speed estimation. A structurally compact and simple neural model is required for real time implementation to derive the desired accuracy and response time. A novel self organizing Single Neuron Cascaded Neural Network (SNC-NN) with moving weight method of learning is proposed in this paper to efficiently model the on-line speed estimator. The proposed neural model is obtained using input/output data and LevenbergMarquardt training algorithm. The performance of the proposed neural model is compared with the popular feed forward architecture in terms of compactness, execution time and accuracy. The SNC-NN based speed estimator is tested through extensive simulations over a wide operating range. The promising results indicate the suitability of the speed estimator for sensor-less induction motor drives.