A method for approximating a specified impulsive response

Given a system\u27s time response, y(t), to a unit impulse, a method is presented by which an approximating function y*(t) may be determined. It is inherent in the method that the transform function Y*(s) is rational. The poles of the approximating function are found from the coefficients of a difference equation, which is made to fit the given data in a least mean square error sense. The zeros are determined by a procedure based on the fact that at a given time the given function and its approximate derivatives are known. Accuracy and implementation of the method are investigated primarily through the medium of example problems. Usefulness and comparison with other methods are among the topics discussed in conclusion --Abstract, page ii