New method to combine conflict evidence based on pignistic probability distance
Yong Sheng Deng · Infrared and Laser Engineering · 2009
Dempster Shafer evidence theory is widely used in data fusion. However, illogical results may be obtained by classical Dempster Shafer combination rule when collected evidence highly conflicts each other. Many methods are proposed to solve this problem. In these methods, the conflict coefficient k is used to measure the conflicting degree between the evidence. However, researches proclaim that not only the conflict coefficient k, but also other relating factors have an effect on conflict degree between evidence. A new combination method based on the classical conflict coefficient k and pignistic probability distance was proposed. A numerical example was used to show that: while the conflict between the evidence is small, the result is close to D-S evidence combination rule; while it is large, the method is more perfect in the rule model due to the more reasonable results.