
Dynamic Scheduling of Assembly Line Using Neural Networks for SARIA Industrial Complex - Sudan
Author(s) -
Khalid M. M. A. Bukkur,
Khalil Ahmed,
Mohammed Shukri
Publication year - 2022
Language(s) - English
DOI - 10.54388/jkues.v2i1.4
Subject(s) - artificial neural network , scheduling (production processes) , computer science , robustness (evolution) , schedule , dynamic priority scheduling , agile software development , industrial engineering , real time computing , mathematical optimization , engineering , artificial intelligence , mathematics , biochemistry , chemistry , software engineering , gene , operating system
In this paper, the neural network was developed to improve the dynamic scheduling of SARIA industry complex. This work implemented into two modules: modeling system by time calculation, the main purpose of this modeling is to calculate the total manufacturing times of the products. The second module is the neural network model architecture, constructed to hold a real-time optimization schedule to solve dynamic scheduling problems. The analytical model was built, including collection and manipulation of data, time calculations, and the neural networks model was formulated. Several training tests were carried out, then the dynamic scheduling was formulated. To assess the validity of the system and to investigate the efficiency and robustness of the system, the results were compared with those obtained from SARIA. The results reveal that the total time of products demand is easily calculated, and the system is agile to schedule any change that occurs in the demand, also the proposed system reduces 4 shift days for one demand. So the developed neural network leads to minimizing the total costs.