Predictive Modeling of Calibration Cycle and Calibration Conditions

Download or read book Predictive Modeling of Calibration Cycle and Calibration Conditions written by Ly Heng Phey (Graduate student) and published by . This book was released on 2021 with total page 83 pages. Available in PDF, EPUB and Kindle. Book excerpt: Abstract: The research focuses on statistical learning applications such as survival analysis and machine learning to an engineering reliability problem. The goal of this research is to apply survival analysis and statistical learning techniques to determine the reliability, the calibration interval, and the calibration condition of the test equipment. The research is intended to develop the appropriate statistical model to predict the calibration interval of different test equipment. Methods to be considered include extended cox-proportional model and many machine learning techniques such as k-nearest neighbors, logistics regression, and support vector machines. Currently, parametric analysis such as an exponential model is implemented to determine the reliability and calibration interval. The disadvantage of this exponential model is that it is strictly based on pass/fail and resubmission time. This model does not consider the potential risk factors that might affect the true performance of the test equipment. Since different test equipment has different characteristics and purposes, it is important to include the interest risk factors/covariates into the reliability function to predict the most appropriate calibration interval (calibration cycle). Thanks to the results of this research, interval and reliability analyst now have an ability to predict/observe the reliability of the test equipment throughout the equipment life cycle.