Estimation of the measurement uncertainty based on quasi Monte-Carlo method in optical measurement

Because measurement uncertainty is an important parameter to evaluate the reliability of measurement results, it is essential to present reliable methods to evaluate the measurement uncertainty especially in precise optical measurement. Though Monte-Carlo (MC) method has been applied to estimate the measurement uncertainty in recent years, this method, however, has some shortcomings such as low convergence and unstable results. Therefore its' application is limited. To evaluate the measurement uncertainty in a fast and robust way, Quasi Monte-Carlo (QMC) method is adopted in this paper. In the estimating process, more homogeneous random numbers (quasi random numbers) are generated based on Halton's sequence, and then these random numbers are transformed into the desired distribution random numbers. An experiment of cylinder measurement is given. The results show that the Quasi Monte-Carlo method has higher convergence rate and more stable evaluation results than that of Monte-Carlo method. Therefore, the quasi Monte-Carlo method can be applied efficiently to evaluate the measurement uncertainty.