Modeling of Machining Errors’ Accumulation Driven by RFID Graphical Deduction Computing in Multistage Machining Processes

Abstract

Interpreting the mechanism of machining errors' accumulation (MEA) is hard in multistage machining processes. This article addresses a novel three-dimensional graphical model named RFID-MEA for quantitating MEA inspired by the radio-frequency identification (RFID) graphical deduction computing model. RFID-MEA aims at revealing the MEA's mechanism instead regarding the machining process as a traditional “black box” model. First, Taylor's second-order expansion is employed to advise the mathematical representation of MEA and its calculation formula. Then, RFID data are collected to bond the MEA formula's variables with certain data silos in the right type at the right time. After that, raw not only structured query language (NoSQL) data computing is executed in edge nodes for assigning values to variables of the MEA formula accurately. Finally, machining errors are estimated by solving the MEA formula. By illustrating an industrial case, it is shown that the RFID-MEA has better performance in estimating machining errors, as well as with higher reliability and interpretability.