基于改进的ε-滤波LMS算法的一种自适应逆控制系统

A Modified Filtered-ε LMS Algorithm Based Adaptive Inverse Control System

针对导弹在包线内飞行时,要求其自动驾驶仪对于飞行速度和飞行高度的变化表现出良好的一致跟踪性能,文中采用自适应逆控制(Adaptive Inverse Control)结构设计了某型自动驾驶仪的过载通道,以使自动驾驶仪的输出过载正确地跟踪指令输入。基于ε-滤波的 LMS(Least Mean Square)算法被证明是一种稳定、有效的高速自适应过程,在自适应控制系统中得到了广泛的应用。文中针对基于ε-滤波的 LMS 算法的自适应逆控制系统中,需要进行对象模型及其逆模型的建模问题,提出了一种近似建模方法,简化了系统设计。仿真结果表明,这种基于改进的ε-滤波的 LMS 算法的导弹自动驾驶仪自适应逆控制系统能够很好地跟踪指令输入,并不受飞行速度和飞行高度变化的影响,表明了其在导弹飞行包线内实施控制的有效性。

Good tracking performance is required when designing an autopilot of a missile,due to the dramatic changes of flight altitude and Mach numbers throughout the flight envelope.An adaptive inverse control structure is adopted to design the normal acceleration channel of an autopilot,so that the autopilot can track the command input accuratdy.Filtered-ε LMS algorithm has been proven to be a stable,efficient and fast adaptive process,and has been applied in many adaptive control systems,but plant model and inverse plant model are needed in this algorithm.In order to simplify the system design,an approximate model- ing method is presented in this paper.Simulation shows the modified filtered-ε LMS algorithm based adaptive inverse control of an autopilot of a missile works very well.