一种新型化学溶液复介电常数测量系统设计  

Design of a new system for measuring the complex permittivity of chemical solution

作　　者：胡自立[1] 陈倩[1] 黄卡玛[1] 杨阳[1] 

机构地区：[1]四川大学电子信息学院,四川成都610065

出　　处：《化学研究与应用》2017年第5期680-684,共5页Chemical Research and Application

基　　金：国家自然科学基金项目(61501311)资助;国家重点基础研究发展计划(2013CB328902)资助

摘　　要：设计了一种圆柱腔体的测量装置,并与矢量网络分析仪设计加工成一个整体。利用FDTD(时域有限差分)与神经网络算法结合反演计算化学溶液的介电常数。将介电常数的实部和损耗角正切放在同一网络中计算,在增加计算难度的基础上极大的增加了计算的准确度。利用Matlab编程实现装置与电脑端的串口数据通信,测量数据的神经网络计算,以及最终人机交互GUI(图形用户界面)设计。系统设计为一个整体,操作简单,实验结果表明,实部最大相对误差不超过10%,测量结果在合理的范围内。A cylindrical cavity permittivity measurement is designed, and it was designed and made into a whole with the vector net-work analyze. The FDTD （finite-difference time-domain） and neural network algorithm is used to calculation the dielectric constant of chemical solution. The real part and loss angle tangent of permittivity is put into a same net,then it increases the difficulties and accuracy of calculation. The Matlab is used to realize the serial data communication of device and computer,neural network calcula-tion of the measured data,and the design of man-machine interface. The system is designed as a whole so it is easy to operate,and the result of experiment shows that the maximum relative error of the true part is less than 10%,and the measurement results are in reasonable range.

关 键 词：介电常数 MATLAB 圆柱腔体 神经网络 

分 类 号：O441.5[理学—电磁学] TM931[理学—物理]