论文信息 - Optimization of Saprolite Ore Composites Reduction Process Using Artificial Neural Network (ANN)

Optimization of Saprolite Ore Composites Reduction Process Using Artificial Neural Network (ANN)

Abstract Indonesia is the third largest country that has laterite reserves. The potential for resources and nickel ore reserves is quite large in Indonesia, but nickel content in nature is very small. One type of laterite nickel ore is saprolite ore, which has small Fe and large Ni content (around 1.5 - 2.5%). In accordance with the Law No. 4 of 2009 concerning Mineral and Coal Mining in increasing the added value of nickel ore, it is necessary to process and refine nickel ore. One of the nickel ore processing/refining technologies is through Pyrometallurgy technique. Pyrometallurgy technique involves high temperature and large energy. The reduction process is one of the nickel ore processing process using the Pyrometallurgy technique. In addition to the reduction process, the use of composites which are mixing of saprolite ore, coal, additive and bentonite has an important role. There are several factors that influence the reduction process in saprolite ore composites. The results of this reduction process are analyzed using X Ray-Difference Fluorescence (XRF). The objective of this research is to obtain an optimal factor combination of the reduction process of saprolite ore composites, which is important to develop effective, efficient and systematic methods. This study utilises a neural network approach that will produce optimal factors with the estimate on the composition in the reduction process of saprolite ore composites. The optimal factor combination is a coal ratio of 15% with a type of additive Ca2SO4 or Composite SB15Ca10P2 with a temperature of 1200 0C and time duration of 3 hours.

[1] Riyanarto Sarno,et al. Estimating Gas Concentration using Artificial Neural Network for Electronic Nose , 2017 .

[2] T. Sun,et al. Effect of coal levels during direct reduction roasting of high phosphorus oolitic hematite ore in a tunnel kiln , 2012 .

[3] Fei Han,et al. A survey on metaheuristic optimization for random single-hidden layer feedforward neural network , 2019, Neurocomputing.

[4] Abhijit Majumdar. Comparative Analysis of Regression and Artificial Neural Network Models for the Prediction of Yarn Hairiness , 2010 .

[5] J. Soedarsono,et al. Effect of Reduction Process Parameter in Direct Reduction Process of Laterite to Produce Substitute Pig Iron for Thin Wall Ductile Iron Material , 2014 .

[6] S. Harjanto,et al. Reduction of Composite Pellet Containing Indonesia Lateritic Iron Ore as Raw Material for Producing TWDI , 2013 .

[7] Safwan Altarazi,et al. Artificial neural network modeling to evaluate polyvinylchloride composites’ properties , 2018, Computational Materials Science.

[8] R. Chaurasia,et al. Application of artificial neural networks and response surface methodology approaches for the prediction of oil agglomeration process , 2018 .

[9] Yue Zhang,et al. An optimization method for multi-objective and multi-factor designing of a ceramic slurry: Combining orthogonal experimental design with artificial neural networks , 2018, Ceramics International.