A BP Neural Network Realization in the Measurement of Material Permittivity

Effective complex permittivity measurements of materials are important in microwave engineering and microwave chemistry. The BP (Back Propagation) neural network computational module has been applied to microwave technology and becomes a useful tool recently. A neural network can be trained to learn the behavior of an effective complex permittivity of material under microwave irradiation in an experimental system. It can provide a fast and accurate result for the material permittivity. Thus, the on-line measurement has been realized. In this paper, a measurement system has been designed and the S-parameters are obtained by full-wave simulations to reconstruct the material permittivity. Moreover, several organic solvents have been measured. The relative errors of the reconstructed results for several organic solvents are less than 5% compared with reference data. The reconstructed results of the effective permittivities of solvents by means of the BP neural network are obtained quickly and accurately.