A Method for Interval-Bound Estimation Based on Grey Theory

Interval analysis is an important measure for uncertainty processing, especially in the case that the sample is small or relevant information is lack. The determination of bounds of uncertain parameters is the precondition to do interval analysis. As traditional Bootstrap method has the weakness that the estimated interval bounds are always too narrow, a new method for interval-bound estimation based on grey theory is proposed. A numerical example about reliability calculation is illustrated to validate the effectiveness of the proposed method, and evidence theory is used to calculate the system reliability. The results show the effectiveness and correction of the proposed method.