针对一类非线性系统 ,提出了一种综合集成滑模控制器设计方法 .首先 ,采用遗传算法进行参数调节 ,从而构造出一个最佳切换函数 ,使得系统既具有良好的动态性能 ,又扩大了鲁棒区域 .然后 ,采用神经网络在线调整控制器参数 ,从而克服了由不确定性引起的系统轨迹偏离切换函数 ,并且通过变学习率学习算法来加快神经网络的收敛性 .最后 ,研究了所提出方法在凹版印刷机张力调节中的应用 .仿真结果表明该控制器具有较好的动态性能及鲁棒性 ,并有效减少了抖动 . Aiming at a class of nonlinear systems, an integrated sliding mode controller design method is proposed.Firstly, genetic algorithm is used to adjust the parameters to construct an optimal switching function, which makes the system not only have good dynamic performance but also expand Robust region.Secondly, neural network is used to adjust the controller parameters online to overcome the uncertainty of the system trajectory deviation switching function, and through the learning rate learning algorithm to speed up the convergence of neural network.Finally, The application of the method to the tension control of intaglio printing machine is presented.The simulation results show that the controller has good dynamic performance and robustness and reduces jitter effectively.