A Speed Estimation Method with Stator Resistance Online Identification based on MRAS

A speed estimation method with stator resistance online identification based on model reference adaptive system is designed for speed sensorless control of induction motor in order to improve the poor control accuracy caused by stator resistance variation under the long-running conditions. For the stator resistance identification, the error between the measured value of stator current and the estimated value of current model is back propagated to adjust the weights of the neural network. In model reference adaptive system, PI control is replaced with signal neuron control to enhance the robustness of system. Simulation results have shown that the improved speed estimation with stator resistance identification has better accuracy and stronger stability compared with general MRAS sensorless control.