In order to solve the problem of parameter estimation in the high-performance control of squirrel cage asynchronous motor，a method for joint parameter identification of asynchronous motors with dual models based on improved whale algorithm was proposed. This method can effectively identify the stator resistance，the rotor resistance，mutual inductance and leakage inductance. In order to improve the identification accuracy of the algorithm，the nonlinear convergence factor was adopted，and the ideas of chaotic reverse learning，simulated annealing and adaptive mutation perturbation were integrated to overcome the shortcomings of the whale algorithm，which relied on the initial population，was easy to fall into local optimum，and had low convergence accuracy. Moreover，combining the advantages of the two traditional motor models，an improved dual-model joint identification was proposed，which further improves the accuracy of parameter identification. Based on this model，the improved whale algorithm was compared with the other two algorithms for motor parameter identification，and the experimental results show that the improved algorithm has high recognition accuracy，which proves the feasibility of applying the algorithm to identify the parameters of the squirrel cage asynchronous motor，and is of great significance for improving the control performance of the squirrel cage asynchronous motor.