A numerical approach to find parameters using experimental data
Ihtzaz Qamar, Syed Wilayat Husain · Communications in Applied Numerical Methods · 1991
Abstract A numerical technique is presented to determine the unknown parameters in a given relationship using the experimental data. In this technique we formulate the required number of independent relationships by integration applied to various parts of the data along with the criterion of the least‐squares error with respect to one parameter. The results for two sample systems, one with two parameters and the other with three parameters, are presented. Comparison with the published work shows that the proposed method is less sensitive to the inaccuracies in the experimental data and generates a better fit than the existing methods such as the ordinary least‐squares technique. The method can easily be extended to systems with any number of parameters.