基于RBF网络的航空发动机 terminal滑模控制

针对现代航空发动机是一个具有不确定性的强非线性系统,结合滑模控制和神经网络控制的优点,提出了一种基于径向基函数(radical basis function,简称RBF)网络的航空发动机terminal滑模控制方法.分析了传统指数趋近律的不足,提出了一种改进的指数趋近律来削弱抖振.该控制器采用terminal滑模面,并且利用径向基函数神经网络在线实时补偿未知干扰和不确定项的影响.仿真结果表明,所设计的控制器取得了令人满意的控制效果,能有效地抑制干扰和参数不确定性的影响,削弱了抖振. 

Terminal sliding mode controller for aero-engine based on RBF neural network

A method of terminal sliding mode control was put forward for aero-engine with uncertainty and strong nonlinearity based on radical basis function (RBF) neural network.The method integrated neural network control into sliding mode control.Shortage of traditional exponent reaching law was analyzed,and an improved exponent reaching law was proposed to weaken the chattering.Controller was devised using terminal sliding mode surface.And influence caused by unknown disturbance and uncertainty was compensated by RBF neural network in a real-time manner.Simulation results show that the devised controller has good effect,effectively restrains the influence and weakens the chattering.