Submit Manuscript  

Article Details


Modelling to Predict Moisture Ratio in Infrared Drying of Machine Plaster by Particle Swarm Optimization

Author(s):

Mehmet Kalender*, Mahmut Temel Özdemir and Hasan Güler   Pages 1 - 10 ( 10 )

Abstract:


Background: Gypsum plaster is one of the most important building materials. Use of gypsum plasters are very common due to their many advantages. Drying process is an important stage in production of gypsum materials and applications. Modelling of drying phenomenon can benefit the drying technology. Recently, particle swarm optimization (PSO) technique has been used to obtain optimum model equations for drying processes.

Objective: The aim of this study is to determine a new modeling approach to infrared drying of machine plaster by using PSO.

Methods: Experimental studies supplied by previous work in the literature have been performed by a laboratory scale infrared dryer in the temperature range 50-70 °C and at atmospheric condition. Experimental moisture ratio values were compared with various mathematical model equations developed for drying process by using particle swarm optimization (PSO) technique.

Results: Fitting tests indicate that the results obtained from PSO technique is better than those of the previous study because of lower χ2, RMSE, and RSS values. The best model equation was the model based on Newton drying equation existing in previous study. However, the model equation derived by Modified Page has been determined as the most compatible model with the experimental data.

Conclusion: It can be said that PSO is successively and reliably used to predict or optimize the experimental data of drying phenomena.

Keywords:

PSO, infrared drying, machine plaster, modeling

Affiliation:

Department of Bioengineering, Fırat University, 23100, Elazığ, Department of Electrical and Electronics Engineering, Engineering Faculty, Fırat University, 23100, Elazığ, Department of Electrical and Electronics Engineering, Engineering Faculty, Fırat University, 23100, Elazığ



Read Full-Text article