关键词: 氢原子频标, 比例-积分-微分控制系统, 反向传播神经网络

Abstract: The stable operation of the hydrogen atomic frequency standard is the prerequisite for its important role,and the role of the Proportion Integration Differentiation (PID) control system is to adjust the output frequency of the crystal oscillator part of the hydrogen atomic frequency standard as close to its nominal value as possible,which is an important part of ensuring the stable operation of the hydrogen atomic frequency standard.Although the parameters of the PID control system will not change over time,their values are still mainly adjusted manually or decisions are made based on accumulated knowledge and a large amount of experimental data.Therefore,this paper adopts a new intelligent technology based on the Back Propagation (BP) neural network.The core of this algorithm is feedback adjustment based on the deviation between the output of the PID control system and the expected output.That is,in the back propagation process,feedback functions are used to better adapt to the situation in actual applications,and they are used as learning objects,so that the best performance can be achieved without external interference and the possibility and scope of deviation from expectations can be effectively reduced.The optimization of parameters in the PID control system is achieved through signal forward propagation and error reverse transmission.The final simulation result is that the daily frequency stability of the hydrogen atomic frequency standard is 1.51E-15.After experimental comparison,this method obtained optimal simulation data,thereby improving the stability of the hydrogen atomic frequency standard operation.

Key words: Hydrogen atomic frequency standard, Proportion Integration Differentiation（PID） control system, Back Propagation (BP) neural network