A Multi-objective Programming Method Based on Relative Deviation

Fachao Li,Kaixin Feng

DOI: https://doi.org/10.1109/liss.2018.8593236

2018-01-01

Abstract:Multi-objective programming, as an effective mathematical model for dealing with multiple objectives, is widely used in the fields of complex system optimization, management decision analysis and artificial intelligence. This paper first analyzes the characteristics and deficiencies of the multi-objective planning model based on priority factors, and proposes a multi-objective planning method based on relative deviations. It mainly introduces the relative deviation to describe the degree of deviation from the target value and reflect the satisfactory state of the decision results. It establishes a planning model in which each objective is at the same level, and analyzes the effectiveness of the method in conjunction with a specific case. The comprehensive analysis and the corresponding conclusions show that the solution model established not only ensures the existence of the solution, but also guarantees the reliability of the decision-making result. It has strong explanatory and operability, and to a certain extent, enriches the existing fuzzy decision-making method.