Risk Analysis by a Probabilistic Model of the Measurement Process

The aim of the article is presentation of the testing methodology and results of examination the probabilistic model of the measurement process. The case study concerns the determination of the risk of an incorrect decision in the assessment of the compliance of products by measurement. Measurand is characterized by the generalized Rayleigh distribution. The model of the measurement process was tested in parallel mode by six risk metrics. An undesirable effect in the reconstruction building block of the model was detected, consisting in the distortion of probability distribution at the edges of the measuring range. The paper gives guidelines on how to use the model, to obtain the analytical risk assessment consistent with the results of the Monte Carlo method. The study can be useful in product quality control, test design, and fault diagnosis.

Automated summary

The article presents the testing methodology and results of examining the probabilistic model of the measurement process using six risk metrics. An undesirable effect of probability distribution distortion at the edges of the measuring range was detected. Guidelines on using the model to obtain consistent risk assessment with Monte Carlo method results are provided.