Objective Aiming at the problem of nonlinear instability and hysteresis of the magnetic levitation ball system an adaptive gray prediction composite control strategy based on particle swarm optimization was proposed. Methods A grey predictor with equal-dimension and new-info characteristics was introduced into the feedback loop of the PID control module to provide timely feedback correction of system errors so as to improve the response speed and robustness of the control system. And the particle swarm intelligence algorithm was integrated to iteratively optimize the controller parameters so as to improve the control accuracy and anti-interference ability of the control system. Finally a simulation platform was constructed in the MATLAB / Simulink environment for comparative experiments. Results The experimental results showed that the overshoot peak time and adjustment time of the adaptive grey predictive control system model based on particle swarm optimization were significantly improved. Conclusion It is confirmed that this strategy can effectively suppress the system hysteresis and has good stability and robustness.