New buffer operators with variable weight based on average tempo and their optimization

Abstract

Purpose: The purpose of this paper is to solve the problem existing in the forecast of impact disturbance grey system.Design/methodology/approach: Under the axiomatic system of buffer operator in grey system theory, a novel kind of buffer operators with variable weight λ based on the principle of average tempo of time sequence and using new information is proposed. The optimization solution for variable weight λ is obtained by using genetic algorithm. It is proved that the new buffer operators are effective.Findings: The results show that the new buffer operators accord with the buffer operator's three axioms and the monotonicity non‐variable axiom. It is proved theoretically and in practice that the new buffer operators are more useful than other buffer operators in grey modeling for sequence with impact disturbance.Practical implications: The novel buffer operators can reduce the randomness of grey sequence distorted by impact factors, and the forecast accuracy of a model which is built through the process with a novel buffer operator is significantly increased.Originality/value: The paper succeeds in constructing two novel buffer operators with variable weight and the properties of novel operators are studied. The method to solve optimization value of weight is proposed. The method widens the scope of grey model application.