Application of K-Means Clustering Algorithm to Weight Measurement System in a Series Production Line

Author :  

Year-Number: 2024-54
Language :
Subject :

Abstract

In this study; A weight measurement system has been developed in order to control the weights of heavy-duty air compressors after assembly and to record the weight data. In particular, air compressors are widely used to produce compressed air in heavy vehicles. Many electronic and mechanical products on the vehicle, especially the brake systems, work with compressed air. The interface was designed and the program was developed to keep the data and process under control in the serial production line compressor weight measurement system. C# programming language and SQL database were used to manage data traceability. Production work order forms form the basis of the weight recording system. Work order numbers prepared according to the production plan are unique and once the work orders are completed, the work order is not created again with the same barcode number. The data was recorded in the SQL database and the weight data of each compressor was matched with the work order barcode number and compressor code. The weight ranges of each compressor were determined with the first 100 compressors using the learning mode. Automatic comparison of weights was calculated in the k-means clustering algorithm and the approval-rejection stage was completed by checking the reference range limits of the compressor. Using the learning mode in the system provides optimum minimum-maximum It provided an advantage in determining the weights. Performance tests are applied to the compressors whose assembly is completed before the weight measurement stage. In addition to performance tests, product weight measurements will prevent possible errors and increase efficiency and quality in production. The purpose of ensuring traceability of data is to diagnose errors that may occur with the creation of a dataset. Resolving errors is one of the main factors in producing products with high quality and efficiency.

Keywords

Abstract

Keywords


                                                                                                                                                                                                        
  • Article Statistics