基于神经网络的AD8302相位响应特性曲线校正方法  

Correction method of AD8302 phase difference response characteristic curve based on neural network

作　　者：李苑[1] 陈光达[1] 

机构地区：[1]西安电子科技大学机械电子工程系,陕西西安710071

出　　处：《信息通信》2017年第4期45-47,共3页Information & Communications

摘　　要：随着无线通信技术的发展及其在民用、军事、国防等领域的广泛应用,天线系统对功率有了更高的要求。在阵列天线系统中,天线单元通过功率合成提高发射功率,相位差是功率合成技术的关键,是科学研究不可缺少的部分。因此,如何获得可靠的相位差是值得研究的课题。基于AD8302芯片的相位差检测方法是常用的相位检测法,其检测频率范围大,检测精度高,硬件电路实现简单,但该芯片在测量某些相位差时,由于特性曲线的问题测量误差较大,在实际应用中造成了很大的困难。所以文章针对该问题设计了一套基于神经网络的处理方法,有效地补偿了AD8302自身的缺陷,使得相位差测量精度在为0.1°,误差约为±0.5°。With the development of wireless communication technology and widely used incivilian, military, national defense etc. Antenna system has a higher demand on the power. In an array antenna system, antenna unit improve the emission power by means of power synthesis. However, phase difference is the key to power synthesis technology, is an indispensable part of scientific research. Therefore, how to obtain a reliable phase difference deserves to be researched. Phase difference detection method based on AD8302 chip is a common phase assay,with the test frequency range,high accuracy,and simple hardware cir- cuit implementation.But the chips have many shortcomings make certain phase measurement error is bigger, caused a lot of dif-ficulties in the practical application. So this paper aimed at the defect of AD8302 design a set of processing method based on neural network, effectively compensate of the flaws of AD8302 makes phase difference measurement is 0.1 degree, and the error is about 0.5.

关 键 词：测量与计量 相位差 AD8302 神经网络 

分 类 号：TN874[电子电信—信息与通信工程]