Because of time, cost or safety restrictions, multi-sensor measurement results are commonly
of poor information characterized by small data samples and an unknown data distribution.
A dynamic bootstrap grey method is proposed to handle such problems considering that traditional
statistical methods cannot. For small data samples, the proposed method has a
lower relative estimation error of the measurement results compared to the grey bootstrap
method and the Monte Carlo method. It is also superior to the Bessel method for large data
samples. Based on two sets of experimental data, the estimation reliability of the dynamic
bootstrap grey method is above 95% with a confidence level of 99.7% while providing a very
low relative error of the estimated expected measured value and estimated measurement
uncertainty. The presented results show that the dynamic bootstrap grey method can estimate
the measurement results successfully without requiring data distribution information
or a large sample size.