Skip to Main Content
Series of real experiments for obtaining of reliable data of displacement parameters of a plane object under testing by the microwave interference method using neural network technologies have been carried out. Two mathematical models of neural networks were generated. The first model was created for obtaining the quadratic law of a detector characteristic; the second one was for updating estimations of displacement parameters. The neural network application has reduced the error of estimations.
Date of Conference: 8-12 Sept. 2008