为提高导引头预定回路的动态性能,提出一种基于Hebb规则的改进单神经元PID控制算法。首先,将传统PID控制与单神经元学习规则相结合,从而实现了参数的在线调节,获得了更好的动态性能;其次,选择误差平方作为优化指标,并在负梯度搜索算法中引入惯性项,设计了基于Hebb规则的改进单神经元PID控制算法。仿真结果表明,改进的控制算法在快速性方面优于常规的单神经元PID控制,并且有较好的鲁棒性和抗干扰性。

In order to improve the performance of seeker scan loop, a modified single neuron PID controller was put forward based on the rule of Hebb. Firstly, a single neuron PID controller was designed to achieve online adjustment of PID parameters on the basis of conventional PID controller and neuron studying regulation with better performance. Secondly, with the square of error selected as the performance index, a single neuron PID control algorithm was developed by means of negative gradient algorithm which contains an inertia part. Simulation results show that the modified algorithm has the advantage of faster response over conventional single neuron PID controller as well as the merits of robustness and anti-disturbance.