Effects of unequal inputoutput sampling rates on the identification of electromagnetictransient systems
Abstract
Some limitations on the identification of electromagnetic systems based on input and output data of different record lengths and sampling rates are discussed. This problem has come to attention when trying to process internal data for the A7E A7E scale model in the Large Scale Model Measurement Program. The problems arise due to limitations of the measurement equipment and computer algorithms used to process the data. Because the input and output records must have the same sampling rates to be processed, considerable data preprocessing must be done to ensure that this condition exists. Three such preprocessing methods are included for comparison, demonstrated with examples, and the second one is concluded to be preferred for general use. Included are examples of the identification of a system model for each of the above three cases, along with a brief discussion of each. Brief mention is made of the importance of proper windowing of truncated data sequences in order to obtain correct Fourier transforms using the discrete Fourier transform. Examples demonstrate the importance of windowing properly.
 Publication:

NASA STI/Recon Technical Report N
 Pub Date:
 August 1982
 Bibcode:
 1982STIN...8324762B
 Keywords:

 Algorithms;
 Data Sampling;
 Electromagnetism;
 Fourier Transformation;
 Input/Output Routines;
 Mathematical Models;
 Analysis (Mathematics);
 Computer Programs;
 Computerized Simulation;
 Electromagnetic Measurement;
 Sampling;
 Communications and Radar