Efficient Methods for Evaluating Task-Specific Uncertainty in Laser-Tracking Measurement

Abstract

For task-specific coordinate metrology, it is necessary to associate with task-specific parameters an estimate of their uncertainty at a specified level of confidence. In recent years, various sources of uncertainty for laser trackers themselves have been studied. However, methods for determining the uncertainty in laser-tracker-based task-specific measurement must be further studied. In this paper, the guide to the expression of uncertainty in measurement (GUM) uncertainty framework (GUF), the Monte Carlo method and a hybrid method combining grey evaluation and neural network technologies are used to evaluate the task-specific uncertainty in laser-tracking measurement. First, this article discusses the contributions to measurement uncertainty in specific laser-tracking measurement, including instruments themselves, data fusion, measurement strategies, measurement environment and task-specific data processing. Second, the principles of GUF, Monte Carlo and the aforementioned hybrid method are presented. Finally, a case study involving the uncertainty evaluation of a cylindricity measurement process using the above-mentioned methods is illustrated. The results demonstrate that these methods have different characteristics in task-specific uncertainty evaluation.