一种基于神经网络的复介电常数测量装置

物质的复介电常数反映了物质对微波的吸收和透射能力,为了提高微波工业应用中能量的利用效率,能准确测量物质的介电常数就显得尤为重要。本文设计了一种在2.45 GHz下的单端口圆柱腔体测量装置,内置聚四氟乙烯容器放置待测物质,避免化学溶液与腔体的金属发生反应影响测量准确性。使用圆形贴片天线向腔体内馈波,提高物质吸收微波的均匀性;利用单端口的反射参数,使用神经网络对复介电常数进行反演;使用双分支神经网络分别提取介电常数实部和损耗角正切值的特征,改善神经网络的训练效果,提高结果准确度,测试结果表明最大误差不超过10%,反演结果与实际值一致性较好。

A Complex Permittivity Measuring Device Based on Neural Network

The complex permittivity of a substance reflects its ability of microwave absorption and transmission. In order to improve the energy utilization efficiency in microwave industrial application, it is particularly important to accurately measure the permittivity of a substance. A single port cylindrical cavity measuring device at 2.45 GHz is designed, in which a polytetrafluoroethylene container is built to place the measured material, avoiding the reaction between the chemical solution and the metal in the cavity which affects the accuracy of measurement. A circular patch antenna is used to feed wave into the cavity to improve the uniformity of microwave absorption. By using the reflection parameters of single port, the complex permittivity is inversed by neural network. A double branch neural network is used to extract the characteristics of the real part of the dielectric constant and the tangent value of the loss angle, so as to improve the training effect of the neural network and the accuracy of the results. The test results show that the maximum error is less than 10%, and the inversion results are in good agreement with the actual values.