Different experts may give completely conflicting conclusion (i.e.

evidence) for the same issue resulting from the different emphasis and experience

and reliability information fusion for high conflict evidence is a difficult problem in reliability engineering. The existing research focuses on the improved Dempster-Shafer (D-S) evidence theory for solving the present problem

in which an individual expert is allocated to a weight for measuring the authority

importance

and credibility

i.e.

each evidence is assigned by a weight by the users according to the understanding on the experts. The previous improved D-S evidence theory assigns a deterministic constant for each weight. However

it is not easy to give an accurate value for each weight because of the epistemic uncertainties of the user in engineering. Based on the improved D-S evidence theory and possibility theory with interval

the uncertain weight is introduced and a novel fusion method for the problems with high conflict information is proposed

and the possibility theory is employed to overcome the information fusion with high conflict evidence. Finally

an example is employed to demonstrate the feasibility of the present approach.