Title: Moisture content estimation during fixed bed drying process with design of experiment and ANFIS methods

Authors: Mustafa Tahir Akkoyunlu; Engin Pekel; Mehmet Cabir Akkoyunlu; Saban Pusat; Coşkun Özkan; Selin Soner Kara

Addresses: Department of Energy Systems Engineering, Ereğli Faculty of Engineering and Natural Sciences, Necmettin Erbakan University, Ereğli, Konya, Turkey ' Department of Industrial Engineering, Yildiz Technical University, Besiktas, Istanbul, Turkey ' Department of Industrial Engineering, Yildiz Technical University, Besiktas, Istanbul, Turkey ' Department of Mechanical Engineering, Yildiz Technical University, Besiktas, Istanbul, Turkey ' Department of Industrial Engineering, Yildiz Technical University, Besiktas, Istanbul, Turkey ' Department of Industrial Engineering, Yildiz Technical University, Besiktas, Istanbul, Turkey

Abstract: In this study, a two stage methodology was applied to predict the exit coal moisture content during the drying process. The first stage included a design of experiment (DoE) study which made easy to determine the significance levels of drying parameters. At the end of the DoE stage, it was determined that the most significant parameter was bed height, and the least significant parameter was exit air relative humidity. The second stage included an adaptive neuro-fuzzy inference system (ANFIS) method which was applied to estimate the exit coal moisture content at any time. The experimental studies were conducted with different levels of the parameters (air temperature, air velocity, bed height, particle size, and air relative humidity). At the end of the second stage, the applicability of the ANFIS in the estimation of the exit coal moisture content was showed with satisfying results. R2 value was increased from 0.465 to 0.842. [Received: March 28, 2017; Accepted: November 3, 2017]

Keywords: ANFIS; design of experiment; DoE; drying; moisture estimation; low rank coal; LRC; lignite.

DOI: 10.1504/IJOGCT.2019.103066

International Journal of Oil, Gas and Coal Technology, 2019 Vol.22 No.3, pp.332 - 345

Received: 28 Mar 2017
Accepted: 03 Nov 2017
Published online: 15 Oct 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article