A step invariant model based on Gompertz´s function for growth prognosis

Abstract

A step invariant model based on Gompertz´s function is presented. Model adjustment was done on broilers and Eucalyptus grandis growth data. The proposed model presents the form Pf = Pa.exp(b₀.S^b2(exp(b₁.Ia))-(exp(b₁.If))), where Pf refers to the future estimated yield at age (If) based on yield (Pa) observed at present age (Ia). S means the introduction of a new variable to the model, being used as binarian variable to indicate male and female broilers and as a site quality for eucalypt. The main advantages of using the proposed model resides in the compatibility of results, assured by the step invariant characteristics, individuals projections and analysis of errors based on the prognosis intervals.