基于二阶Adaline网络的位标器自适应逆控制方法

为了提高导引头成像传感器视轴在系统参数时变时的稳定度，对基于二阶Adaline网络的位标器自适应逆控制方法进行了研究。先是对位标器的动力学模型和机电模型进行了数学推导与分析，接着通过对Adaline网络学习算法的研究，提出了一种不易受噪声影响的变步长LMS算法用于增强其学习能力。在最后设计的位标器双闭环控制系统中，电流环采用变速PID控制，速度环利用系统输出误差，按改进的二阶Adaline网络算法来实现其自适应逆控制的调节过程。内场实验表明，与PID控制相比，所提出的方法显著提高了位标器系统的控制精度，具有很强的鲁棒稳定性。

Adaptive inverse control method based on second order Adaline network for coordinator

In order to improve the optical axis stability of coordinator seeker imaging sensor while the system parameters was time-varying, an adaptive inverse control method based on second order Adaline network for coordinator seeker was studied. Both of dynamic model and electromechanical model were derived and analyzed by mathematical method firstly, after the study process of second order Adaline network researched, the LMS algorithm which was not easily affected by the noise and adjust the learning step size automatically was proposed, it was used to enhance the learning ability of second order Adaline network. The control system was designed for coordinator seeker adopted double closed-loop control structure, here the current loop used variable PID control method and the speed loop used adaptive inverse control method. The adaptive inverse control process was regulated with system's output error according to the improved second-order Adaline network algorithm. Infield tests show that the control method which was proposed improves the control precision of the coordinator greatly, it has stronger robust stability than PID control.