基于RBF+NTSMC的舰载火箭炮随动系统控制研究

Research on Servo System Control of Shipborne Rocket Gun Based on RBF+NTSMC

针对舰载火箭炮控制系统存在火箭炮射速、跟踪精度的问题,提出了一种基于RBF神经网络的非奇异终端滑模控制(NTSMC)策略。基于非奇异终端滑模控制强鲁棒性的特点,用RBF神经网络对未建模动态进行自适应逼近,将RBF神经网络与非奇异终端滑模控制相结合,既保持系统良好的鲁棒性又提高了火箭炮发射精度。仿真结果表明:与传统的NTSMC相比,该控制策略有效地提高舰载火箭炮系统的响应速度、鲁棒性以及发射的命中精度。

Aiming at the problem of rocket launching speed and tracking accuracy in shipborne rocket gun control system,a non-singular terminal sliding mode control(NTSMC)strategy based on RBF neural network was proposed.Based on the strong robustness of nonsingular terminal sliding mode control,the RBF neural network adaptively approaches the non-modeled dynamics,which can effectively improve the response speed and robustness.The combination of RBF neural network and non-singular terminal sliding mode control not only maintains the system's good robustness but also improves the accuracy of the rocket gun projectile.The simulation results show that compared with the traditional NTSMC,this control strategy can effectively improves the response speed,robustness of shipborne rocket gun system and hitting accuracy of the projectile.