The invention discloses an identifying method for an inertia shaft of a full-automatic magnetic bearing system based on nonlinear self-adaption control. The method comprises the following steps that firstly, a full-automatic magnetic bearing rotor dynamics model with rotor imbalance and displacement sensor harmonic noise is established; secondarily, a nonlinear self-adaption control rule and estimation rule are provided, the system stability and astringency of estimation parameters are proved, it is guaranteed that a magnetic suspension rotor inertia shaft displacement estimation value is convergent to zero, and meanwhile the high-level harmonic component fourier coefficient of displacement sensor harmonic noise is estimated; then, a variable rotation speed strategy is adopted, the same frequency component of displacement sensor harmonic noise and recognizable degree of rotor imbalance are improved, the fourier coefficient values of two same frequency components are solved, and finallyidentifying for the inertia shaft of the full-automatic magnetic bearing system is achieved. According to the identifying method for the inertia shaft of the full-automatic magnetic bearing system based on nonlinear self-adaption control, speed raising or speed reduction for one time is only needed, and identifying of the magnetic bearing inertia shaft can be achieved online, so that a magnetic suspension rotor rotates around the inertia shaft.