Horizontal Accuracy Monitoring System for Precision Machinery
Journal
2023 IEEE 5th Eurasia Conference on IOT, Communication and Engineering, ECICE 2023
ISBN
9798350314694
Date Issued
2023-01-01
Author(s)
Abstract
Horizontal accuracy refers to a crucial factor affecting machining precision. Poor horizontal accuracy of the machine can lead to uneven vertical loading on the feet, causing deformation of the machine bed due to gravity. This resultant error affects the motion straightness and angular accuracy of the equipment, leading to subpar machining flatness. Adjusting the horizontal accuracy of the equipment is achieved by adjusting the heights of the leveling bolts. Currently, most research focused on offline systems for assessing horizontal accuracy with fewer efforts on real-time monitoring mechanisms. Thus, based on precision equipment and a case study of base leveling assembly, we proposed a real-time horizontal accuracy monitoring system. The system included smart feet with force-sensing capabilities and a computer application. The system was used to observe variations in vertical loads on the feet and calculate the horizontal state of the target plane. It offered an interface for users to check real-time foot load values. The relationship between vertical loads and the test platform's tilting angle was constructed by an artificial intelligence model. The user interface displayed the current load, calculated horizontal state, and the standard load under the equipment's horizontal state. Initially, CAD software was used as an assist while reference loads for each foot on the equipment. Then, through experimentation, standard loads of each foot at a horizontal state were determined. These standard load data were the basis for setting horizontal accuracy thresholds. Reference loads as theoretical values were used to verify the reasonableness of the standard loads. In this study, the standard loads for the base leveling assembly process of a precision device were determined using intelligent feet. The real-time loads and device status during this process were displayed on the user interface, aiding assemblers in level alignment. Finally, an online monitoring mechanism was implemented by setting allowable values based on the standard loads. The accuracy of classification for single-axis tilting angle interval in the range of 1 to 1.5 mm/m reached 70%.
Subjects
artificial intelligence | horizontal accuracy | machine tool | real-time monitoring
Type
conference paper