[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

Journal of Information Science and Engineering, Vol. 37 No. 6, pp. 1303-1312

Designing of Artificial Intelligence Methods for Manufacturing Industries Quality Control

Department of Industrial Engineering
German Jordanian University
Amman, 11180 Jordan
E-mail: alab9819@gmail.com; abdallah.abdallah@gju.edu.jo

There is good amount of potential for implementing ANN tool for quality control in manufacturing. In any manufacturing process, there are many input parameters, which are responsible for introduction of variability in the product. If this exceeds acceptance limits specified in the drawing, the component is rejected. However, within acceptance limit guided by the dimensions, tolerances, surface finish or other specifications the variabilty exists, governing normal distribution. The objective of proposed work is to study such process variables, identify their extent of contribution in product acceptance, Design & Develop ANN network model, for such application. Train the network by using training sets defined by domain expert. Validate the results and compare the same by using suitable statistical tools and analysis of the findings for such typical application. Selection of machining parameters is an important task for a specific component. A process engineer (domain expert) who traditionally perform this task manually applies the knowledge that he acquired by learning the mapping between input patterns, consisting of feature being machined (Such as hole, external step) and attributes like (size, tolerance, surface finish etc.) of the part and output pattern, consisting of machining operations to apply to these parts.

Keywords: artificial intelligence, manufacturing, quality control, neural networks, orthogonal arrays

  Retrieve PDF document (JISE_202106_05.pdf)