Estimation of coal moisture content in convective drying process using ANFIS

Abstract

In this study, a new methodology was applied to estimate the coal moisture content during the drying process. Adaptive-network-based fuzzy inference system (ANFIS) was applied to predict the coal moisture content at any time during the drying process. The experiments were carried out for different drying air temperatures (70, 100, 130 and 160°C), drying air velocities (0.4, 0.7 and 1.1m/s), bed heights (80, 130 and 150mm) and sample sizes (20, 35 and 50mm), and the experimental results were used to validate applicability of the ANFIS in the coal drying process. The ANFIS network achieves quite satisfying scientific results with acceptable deviations. The MSE and R2 values were calculated as 1.899 and 0.998, respectively, for the testing stage. The results of this study show the applicability of the ANFIS in the coal drying processes to predict the coal moisture content at any time. Therefore, it is not necessary to carry out all the experiments: by using the ANFIS, the drying curves of some other cases which are not performed can be estimated easily. Herewith, the necessary number of the experiments decreases.