Application of the mind-evolution-based machine learning in mixture-ratio calculation of raw materials cement

Keming Xie, DU Yong-gui, Chengyi Sun

2002 · 23 citations · 0 references

Concepts

Abstract

Mind-evolution-based machine learning (MEBML) is an evolutionary computing algorithm. It inherits "colony" and "evolution" of evolutionism. MEBML adopts "similartaxis" and "dissimilation" operators, which possess the more rapid convergence and the higher calculation accuracy. Aiming at a difficult problem for the accurate mixture ratio of raw materials of cement processing, MEBML is proposed, which calculates the mixture ratio of raw materials of cement and overcomes the defects of general calculation methods. The simulation example is given to show that this algorithm not only has a rapid convergence rate and high calculation accuracy, but also has not the prematurity problem of a genetic algorithm. This algorithm can be applied to any mixture ratio calculation of other materials.