随动系统的新型非奇异快速终端滑模控制

鉴于某舰载随动系统存在摩擦、海况等扰动因素，为提升系统控制精度及鲁棒性，提出了一种基于相位补偿的非奇异快速终端滑模控制（NFTSM+ARBF+TTD）方法。将泰勒公式与非线性函数fhan相结合，构造跟踪微分器，旨在减小相位延迟。在抑制微分过程噪声的基础上，应用泰勒公式进行相位补偿。设计自适应神经网络（ARBF）逼近扰动项，减小了非奇异快速终端滑模（NFTSM）中的抖振。扰动估计结果表明，与RBF相比，应用ARBF逼近扰动能够有效抑制微分峰值，更接近扰动真实值。半实物仿真实验中，正弦跟踪结果表明NFTSM+ARBF控制策略的跟踪误差比NFTSM减小一半左右，具备良好动态跟踪性能，适用于该交流伺服系统。

Novel NFTSM Control for Servo Systems

The shipborne servo system featured friction and intricate disturbances， which needed to improve control precision and robustness. A NFTSM control method was proposed with phase compensation（NFTSM+ARBF+TTD）. Taylors formula-based tracking differentiator（TTD）was innovated with Taylors formula and nonlinear function fhan. The proposed method could compensate phase delay and acquire desired control command， also suppress noises. By using ARBF to approach the disturbance， the chattering problem in NFTSM could be solved effectively. Experimental results show that adaptive radial basis function（ARBF） may suppress the differential peaks and estimate disturbance accurately and effectively. Semi-physical simulation experiments were conducted to tracking sinusoidal reference. The results show that compared with traditional NFTSM， the tracking errors are reduced twice time by NFTSM+ARBF+TTD method. Experimental results show the proposed method may be applied to shipborne rocket launcher AC servo systems.