基于神经网络的雷达天线稳定平台控制设计  

Research on Control System of Radar Antenna Stability Platform Based on Neural Network

作　　者：董睿 俞竹青[1] 晁令锦 DONG Rui;YU Zhu-qing;CHAO Ling-jin(School of Mechanical Engineering,Changzhou University,Changzhou 213164,China)

机构地区：[1]常州大学机械工程学院,常州213164

出　　处：《自动化与仪表》2021年第12期33-37,84,共6页Automation & Instrumentation

摘　　要：为了提高稳定平台的抗干扰能力,使雷达天线指向稳定保证成像清晰,该文提出一种神经网络PID控制方案,此方案是利用神经网络的非线性映射能力和自学习自适应能力,找到最优的PID三个控制参数,使稳定平台这种非线性系统具有良好的控制效果。通过MATLAB仿真实验证明,神经网络PID控制方案的超调量仅有1.05%,稳态时间只有0.61 s。远远小于PID控制,而其对扰动的响应幅值仅为PID控制的18.6%,与传统的PID控制相比,神经网络PID控制具有更优秀的抗干扰能力,这种控制方案更加适合在复杂环境下工作的稳定平台,有效保证其精度和稳定性。In order to improve the anti-jamming capability of the stable platform,make the radar antenna point stable and ensure clear imaging,a neural network PID control scheme is proposed.This scheme is to use the neural network's nonlinear mapping ability and self-learning ability to find the optimal PID three control parameters make the non-linear system of stable platform have good control effect.The MATLAB simulation experiment proves that the overshoot which is far less than that of PID control,and its response amplitude to disturbance is only 18.6%of PID control,which is different from traditional PID control.Compared with PID control,neural network PID control has better anti-interference ability.This control scheme is more suitable for stable platforms that work in complex environments,effectively ensuring its accuracy and stability.

关 键 词：雷达天线稳定平台 神经网络PID控制 传统PID控制 MATLAB仿真 

分 类 号：TP241.3[自动化与计算机技术—检测技术与自动化装置]