Optimized GM(1,1) Grey Model Based onRomberg Algorithm

摘要

This paper analyzes the reason for occurrence of the GM(1,1) error, from theintegral calculus angle discrete GM(1,1). A kind of method is put forward for making use of theRomberg integral calculus formula and the nearest neighbor interpolation method to constructthe background value of the model, giving a Romberg Integral Coefficient form. It has raisedthe simulation and prediction precision of GM(1,1), and its practicability is demonstrated byactual examples.