A Study on the Rapid Parameter Estimation and the Grey Prediction in Richards Model

Abstract

Richards model is a nonlinear curve with four parameters. Usually, the estimation of parameters in Richard model is complicated; and there is little literature on the gray prediction in Richards model is found. Facing these problems, this paper presents a algorithm consisting of the following steps: First, replacing approximately the original data with an arithmetic sequence to rapidly estimate the four parameters of Richards model; then, using them as the initial values to fit the original data by nonlinear least squares, the optimized parameters of Richards model are obtained. The algorithm along with “Kernel” and “IAGO” principles are used for the prediction of grey Richards model. The results from the experiments show that the above algorithms have good practicability and research value.