Abstract: High temperature absorber is a great obstacle for development of stealth material. Si/C/N nano powder is
a potential candidate of high temperature absorber. In order to optimizer and control the dielectric loss of Si/C/N nano powder, neural network (NN)
model was developed. The output of NN model was found to be best fit with the sample data’s. With the help of the NN model, a series of predictions
about the dielectric loss-nitrogen content were made. Finally, the prediction about the influence of nitrogen content of Si/C/N nano powder and frequency
on the dielectric loss predicted by NN model was identical well with the experimental results, which shows that the model established by NN can mirror
the dielectric loss-results relationships and inner regularity in the high temperature absorber.

Key words: Si/C/N nano powder, dielectric loss, neural network, modeling