According to the problem of asynchronous motor speed control system, this paper proposes a cerebellar model aritculation controller coupled with active disturbance rejection controller (CMAC-ADRC) control algorithms based on adaptive parameter identification. The respective advantages of CMAC and ADRC were combined. And CMAC neural network was used for feedforward control. Its online learning was applied which suppressed overshoot system, enhanced the robustness and dynamic performance of the system. ADRC was used for feedback control which further enhanced the anti jamming capability. The inertia was identified by using model reference adaptive parameter identification technique and ADRC compensation factors were optimized. Taking converter and asynchronous motor as control objects, the simulation was carried out. The simulation results showed that the response amplitude caused by disturbance of control system using CMAC-ADRC based on adaptive identification was 44.57% of the one using first-order optimization ADRC, and 17.69% of the one using CMAC-PD. Meanwhile, the recovery time of disturbance was 50% of the one using first-order optimization ADRC, and 60% of the one using CMAC-PD. Some experiments were finished on the experiment platform based on MCU-CPLD-DSP. The experiment result showed that with CMAC-ADRC, the overshoot, rising time, response amplitude caused by disturbance, and recovery time of disturbance were 45.49%, 53.33%, 71% and 76.47% of the one using first-order optimization ADRC, respectively.