Research on Several Typical Distribution Discrete Data Sampling Methods for Evaluating Measurement Uncertainty Using Monte Carlo Method

Abstract

The Monte Carlo Method (MCM) is a method of evaluating measurement uncertainty by using random sampling of probability distributions for distribution propagation. The discrete sampling of the probability density function (PDF) of the input quantity is a key and difficult step in the Monte Carlo method. This article proposes a probability density function (PDF) discrete sampling method based on EXCEL software by analyzing the probability density functions and their corresponding cumulative probability density function relationships of typical distribution types such as normal distribution, rectangular distribution, triangular distribution, trapezoidal distribution, and arcsine distribution, and develops an application program. The experimental verification comparison results show that the standard deviation of the method used in this paper for sampling discrete data is consistent with the theoretical calculation results, and the sampling of discrete data is efficient and accurate.