In this paper, a new hybrid scheme which combines radial basic function (RBF) neural network with a model following sliding mode control technique to take their common features is used to solve attitude control problem of gliding guide projectile. The attitude kinematics model described by second-order nonlinear uncertain system is divided into two single-input single-output subsystems by considering the nonlinearity as disturbance. To avoid generating high control value, the coupled inputs are kept as one nominal input instead of being included in lumped uncertainties. The uncertainties in the plant are cancelled by an adaptive RBF neural networks estimator, which is designed based on Lyapunov theory. To verify the effectiveness of the proposed control strategy, attitude tracking control experiments are simulated under strong internal and external disturbances.

中文翻译：

---

基于RBF神经滑模技术的自适应滑翔弹姿态跟踪控制器设计

本文采用径向基本函数（RBF）神经网络与模型遵循滑模控制技术相结合的新混合方案来解决滑翔导弹的姿态控制问题。将二阶非线性不确定系统描述的姿态运动学模型分为两个单输入单输出子系统，将非线性视为扰动。为避免产生高控制值，耦合输入被保留为一个标称输入，而不是包含在集总不确定性中。工厂中的不确定性被基于李雅普诺夫理论设计的自适应 RBF 神经网络估计器消除。为了验证所提出的控制策略的有效性，