为了减轻非线性动态系统中未知迟滞(Hysteresis)非线性的不良影响,本文利用理论建模结合神经网络-“灰箱”建模与控制的方法,提出控制方案。由于建立的迟滞非线性Preisach模型精度的差异,用于控制方案,所达到的效果并不理想,因此,用神经网络加以补偿,既克服了模型的失配,又减轻了系统中未建模非线性对象和系统干扰的影响,达到了较好的控制效果。通过仿真试验,证明了所提出方案的有效性。 In order to mitigate the adverse effects of unknown Hysteresis nonlinearity in nonlinear dynamic systems, this paper proposes a control scheme by combining theoretical modeling with neural network - “gray box” modeling and control. Due to the difference of accuracy of hysteresis nonlinear Preisach model, the effect achieved by the control scheme is not satisfactory. Therefore, the compensation by neural network not only overcomes the model mismatch but also reduces the unmodeled Linear objects and system interference, to achieve a better control effect. Simulation results show the effectiveness of the proposed scheme.