Loading [MathJax]/extensions/TeX/ieeemacros.js
Estimation of Machine Parameters in Exponential Serial Lines using Feedforward Neural Networks | IEEE Conference Publication | IEEE Xplore

Estimation of Machine Parameters in Exponential Serial Lines using Feedforward Neural Networks


Abstract:

Model construction plays an essential role in manufacturing system analysis. A good model should be able to capture the main production phenomena with high accuracy, whic...Show More

Abstract:

Model construction plays an essential role in manufacturing system analysis. A good model should be able to capture the main production phenomena with high accuracy, which is critical in supporting decision-making and achieving optimal output. One important step of constructing such a mathematical model is to determine the values of model parameters. Traditionally, this step is usually done by extracting and analyzing data from the physical workstations. This process may require significant efforts and may not be accurate enough after a certain period of time, especially for complex and high-dimensional systems. Taking advantages of new techniques driven by Industry 4.0, high data availability in manufacturing system provides practitioners valuable chances to implement machine learning-based approaches (e.g., data mining, artificial intelligence) to perform model construction. In this paper, we study one of the most commonly used machine learning techniques-feedforward neural network-in estimating model parameter of exponential serial production lines, based on system's key performance metrics (i.e., throughput, work-inprocess). Numerical experiments are carried out and results are presented to demonstrate the estimation accuracy of this new approach.
Date of Conference: 20-21 August 2020
Date Added to IEEE Xplore: 08 October 2020
ISBN Information:

ISSN Information:

Conference Location: Hong Kong, China

References

References is not available for this document.