论文部分内容阅读
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。
In order to improve the stability of the visual sensor axis of the seeker in time-varying system parameters, an adaptive inverse control method based on the second-order Adaline network was studied. Firstly, the dynamic and electromechanical models of the marker are mathematically deduced and analyzed. Then, based on the study of Adaline network learning algorithm, a variable step size LMS algorithm which is not affected by noise is proposed to enhance its learning ability. In the final design of double closed-loop control system, the current loop adopts variable-speed PID control, and the speed loop uses the output error of the system to realize its adaptive inverse control adjustment process according to the improved second-order Adaline network algorithm. The field experiments show that compared with PID control, the proposed method significantly improves the control accuracy of the scaler system and has strong robust stability.