Comparison of GUF and Monte Carlo methods to evaluate task-specific uncertainty in laser tracker measurement

Liu

Yue

et al. 2018

Purpose This paper aims to solve the nonlinear problem in the uncertainty evaluation of the measurement of the positioning repeatability (RP) of industrial robots and provide guidance to restrict the uncertainty of measurement of RP (uRP). Design/methodology/approach Firstly, some uncertain sources existing in the measurement procedure of RP are identified. Secondly, the probability distribution function (PDF) of every source is established on the basis of its measurements. Some spatial combined normal distributions are adopted. Then, a method, based on Monte Carlo method (MCM) and established measurement model, is developed for the estimation of uRP. Thirdly, some tests are developed for the identification and validation of the selected PDFs of uncertain sources. Afterwards, the proposed method is applied for the evaluation and validation of the uRP. Finally, influence analyses of some key factors are proposed for the quantification of their relative contributions to uRP. Findings Results show that the proposed method can reasonably and objectively estimate the uRP of the selected industrial robot, and changes of the industrial robots’ position and the laser trackers measurement are correlated. Additionally, the uRP of the selected industrial robot can be restricted by using the results of its key factors on uRP. Originality/value This paper proposes the spatial combined normal distribution to model the uncertainty of the repeatability of the laser tracker and industrial robot. Meanwhile, the proposed method and influence analyses can be used in estimating and restricting the uRP and thus useful in determining whether the RP of a tested industrial robot meets its requirements.

Section: Introductionmentioning

confidence: 97%

Section: Introductionmentioning

confidence: 99%

Uncertainty analysis and evaluation of measurement of the positioning repeatability for industrial robots

Liu

Yue

et al. 2018

“…But, most of researches show the MCM is much stricter and better than the GUM method because of the nonlinearity of their measurement models. (Santolaria and Ginés, 2013; Yang et al , 2014).…”

Section: Introductionmentioning

confidence: 99%

Uncertainty evaluation of measurement of orientation repeatability for industrial robots

Liang

Liu

et al. 2020

Purpose This paper aims to propose a reasonable method to evaluate uncertainty of measurement of industrial robots’ orientation repeatability and solve the non-linear problem existing in its evaluation procedure. Design/methodology/approach Firstly, a measurement model of orientation repeatability, based on laser tracker, is established. Secondly, some factors, influencing the measurement result of orientation repeatability, are identified, and their probability distribution functions are modelled. Thirdly, based on Monte Carlo method, an uncertainty evaluation model and algorithm of measurement of industrial robot’s orientation repeatability are built. Finally, an industrial robot is taken as the research object to validate the rationality of proposed method. Findings Results show that the measurement model of orientation repeatability of industrial robot is non-linear, and the proposed method can reasonably and objectively estimate uncertainty of measurement of industrial robots’ orientation repeatability. Originality/value This paper, based on Monte Carlo method and experimental work, proposes an uncertainty evaluation method of measurement of industrial robots’ orientation repeatability which can solve the non-linear problem and provide a reasonable and objective evaluation. And the stochastic ellipsoid approach is firstly taken to model the repeatability of laser tracker. Additionally, this research is beneficial to decide whether the orientation repeatability of the industrial robot meets its requirements.

“…FEM has been applied to analyze the effect of certain parameters on the performance of instruments, but not for the effect of uncertain parameters (Wu et al, 2009). However, the MCS method (Yang et al, 2014;Wang et al, 2015), by simulating certainties and uncertainties, is suitable for analyzing the influence of all related factors on instruments and models. Because many certain and uncertain factors exist in a measurement process, the MCS is better to analyze a measurement model.…”

Section: Introductionmentioning

confidence: 99%

A four-point measurement model for evaluating the pose of industrial robot and its influence factor analysis

Wang

Liang

et al. 2017

Purpose This paper aims to propose a four-point measurement model for directly measuring the pose (i.e. position and orientation) of industrial robot and reducing its calculating error and measurement uncertainty. Design/methodology/approach A four-point measurement model is proposed for directly measuring poses of industrial robots. First, this model consists of a position measurement model and an orientation model gotten by the position of spherically mounted reflector (SMR). Second, an influence factor analysis, simulated by Monte Carlo simulation, is performed to investigate the influence of certain factors on the accuracy and uncertainty. Third, comparisons with the common method are carried out to verify the advantage of this model. Finally, a test is carried out for evaluating the repeatability of five poses of an industrial robot. Findings In this paper, results show that the proposed model is better than the three-SMRs model in measurement accuracy, measurement uncertainty and computational efficiency. Moreover, both measurement accuracy and measurement uncertainty can be improved by using the proposed influence laws of its key parameters on the proposed model. Originality/value The proposed model can measure poses of industrial robots directly, accurately and effectively. Additionally, influence laws of key factors on the accuracy and uncertainty of the proposed model are given to provide some guidelines for improving the performance of the proposed model.