Fitting Algorithm of Parameterized Continued Fractions with Keeping Endpoints and Its Application

摘要

In order to overcome fitting uncertainty, A nonlinear rational parameterized fitting algorithm with keeping endpoints was presented through inverse difference-continued fraction algorithm, the algorithm has property of keeping endpoints, parameterization of middle data points selected. The parameters obtained through minimum error function make fitting functions have much flexibility. Experimental results illustrate that the proposed algorithm may reduce degree of continued fraction interpolation function, numbers of iteration and avoiding non-existence on interpolation function of the continued fraction.The fitting values obtained are more accuracy, the computing efficiency is much more improved,the fitting errors are more stable, approximate effect is better, and the algorithm will be widely used in predict etc.