Uncertainty Evaluation in Multistage Assembly Process Based on Enhanced OOPN.

Yubing Huang,Weiping Mou,Wei Dai,Yu Zhao

doi:10.3390/e20030164

Abstract

This study investigated the uncertainty of the multistage assembly process from the viewpoint of a stream of defects in the product assembly process. The vulnerable spots were analyzed and the fluctuations were controlled during this process. An uncertainty evaluation model was developed for the assembly process on the basis of an object-oriented Petri net (OOPN) by replacing its transition function with a fitted defect changing function. The definition of entropy in physics was applied to characterize the uncertainty of the model in evaluating the assembly process. The uncertainty was then measured as the entropy of the semi-Markov chain, which could be used to calculate the uncertainty of a specific subset of places, as well as the entire process. The OOPN model could correspond to the Markov process because its reachable token can be directly mapped to the Markov process. Using the steady-state probability combined with the uncertainty evaluation, the vulnerable spots in the assembly process were identified and a scanning test program was proposed to improve the quality of the assembly process. Finally, this work analyzed the assembly process on the basis of the uncertainty of the assembly structure and the variables of the assembly process. Finally, the case of a certain product assembly process was analyzed to test the advantages of this method.

Highlights

In manufacturing mechanical products, product quality is directly determined by the assembly process quality due to the great variety of components and the complexity of the structures involved.A multistage assembly system, which consists of multiple components that must be finished within a given time [1] is widely used in the automotive, ship-building, and appliance industries
The assembly process is found at the end of the manufacturing cycle; if the quality defects are not eliminated in the assembly process, the product delivery cycle and customer satisfaction are directly affected [2]
An uncertainty evaluation model was developed for the multistage assembly process based on an enhanced object-oriented Petri net (OOPN) and Shannon entropy

Summary

Introduction

Product quality is directly determined by the assembly process quality due to the great variety of components and the complexity of the structures involved.A multistage assembly system, which consists of multiple components that must be finished within a given time [1] is widely used in the automotive, ship-building, and appliance industries. Product quality is directly determined by the assembly process quality due to the great variety of components and the complexity of the structures involved. The assembly process is found at the end of the manufacturing cycle; if the quality defects are not eliminated in the assembly process, the product delivery cycle and customer satisfaction are directly affected [2]. To meet this need, the object of quality control should be shifted from the final product to the process itself. The manufacturing process is mainly controlled and analyzed by building a model that uses the Markov chain or Petri net [3]

Objectives

Findings

Discussion

Conclusion