The efficiency optimization control method based on loss model has the advantages of clear physical meaning and fast optimization speed. However, this method is sensitive to the real-time parameters of the motor in the model. To improve the observation effect, it is necessary to accurately obtain the real-time parameters of the motor in the loss model. By designing a stable adaptive law, the key parameters in the loss model are identified adaptively;In order to improve the observing speed and robustness of the state observer for induction motor, the pole placement of the feedback matrix with iron loss is carried out;In order to improve the performance of the algorithm, an adaptive identification algorithm for key parameters is designed according to the stability law, and the theory is verified by simulation.