A computationally effective method for calculating the exponential fit

Abstract

In numerous practical applications, the problem of fitting an exponential curve to noisy measurement data often arises for identifying or tracking model parameters. In embedded systems, the computational load and speed of the computation of the fit are crucial to minimize costs and to maximize performance. In this paper, a method is developed for the rapid and effective computation of the three parameter exponential regression model. Computer simulations are also presented and the proposed method is compared with some existing techniques.