研究了以车体加速度统计值最优为控制目标的半主动悬挂神经网络自适应控制方法 ,该方法对阻尼器件的响应特性要求低 ,可以降低成本 ,便于实现。横向半主动悬挂模型仿真结果表明 ,该控制方法算法简单 ,能适应线路激扰和车辆参数的变化 The semi-active suspension neural network adaptive control method based on the optimal vehicle acceleration statistics is studied. The method has low requirements on the response characteristics of the damping device, which can reduce the cost and facilitate the realization. The simulation results of transverse semi-active suspension model show that the proposed control method is simple and can adapt to the changes of line disturbance and vehicle parameters