Prediction of gross calorific value of coal based on proximate analysis using multiple linear regression and artificial neural networks

Abstract

Gross calorific value (GCV) of coal was predicted by using as-received basis proximate analysis data. Two main objectives of the study were to develop prediction models for GCV using proximate analysis variables and to reveal the distinct predictors of GCV. Multiple linear regression (MLR) and artifcial neural network (ANN) (multilayer perceptron MLP, general regression neural network GRNN, and radial basis function neural network RBFNN) methods were applied to the developed 11 models created by different combinations of the predictor variables. By conducting 10fold cross-validation, the prediction accuracy of the models has been tested by using R 2 , RM SE , M AE , and M AP E. In this study, for the first time in the literature, for a single dataset, maximum number of coal samples were utilized and GRNN and RBFNN methods were used in GCV prediction based on proximate analysis. The results showed that moisture and ash are the most discriminative predictors of GCV and the developed RBFNN-based models produce high performance for GCV prediction. Additionally, performances of the regression methods, from the best to the worst, were RBFNN, GRNN, MLP, and MLR.

Özelli̇k Seçi̇mi̇ İle Bi̇rleşti̇ri̇lmi̇ş Destek Vektör Maki̇neleri̇ni̇ Kullanarak Kömürün Üst Isil Değeri̇ni̇n Kisa Ve Elementel Anali̇z Deği̇şkenleri̇nden Tahmi̇ni̇

“…MLP is a well-known and most commonly used ANN method. In addition, RBFN is one of the other popular ANN methods and has been used successfully in the GCV prediction of coals from the proximate analysis results [18].…”

Yıkanmış Türk Linyit Kömürlerinin Üst Isıl Değerinin Destek Vektör Regresyonu ile Tahmini

“…Ayrıca, kömürün GCV'sini tahmin etmek için farklı veri kümeleri ve veri modelleri ile SVR'yi kullanan bazı çalışmalar da bulunmaktadır (Feng et al, 2015;Fu, 2016;Hadavandi, Hower, & Chelgani, 2017;Qi, Luo, Wei, & Fu, 2019;Tan et al, 2015). Diğer yandan, ANN'ler farklı alanlarda gerçek dünyada problemlerini çözmek için yaygın olarak kullanılmaktadır (Açıkkar & Sivrikaya, 2018b;Huang, Liu, & Ren, 2018;Ren & Gao, 2011). Bu çalışmada daha önce ifade edildiği üzere SVR ile karşılaştırma yapmak amacıyla GCV tahmininde üç tip ANN, yani MLP, GRNN ve RBFN kullanılmıştır.…”

“…Bir giriş katmanından, bir veya daha fazla gizli katmandan ve bir çıkış katmanından oluşan MLP, ANN yöntemleri arasında esnek, genel amaçlı ve en çok kullanılan bir ağdır (Açıkkar & Sivrikaya, 2018b). Her gizli katmandaki en uygun nöron sayısını bulmak, MLP'nin çözümlenmesi gereken en önemli konularından biridir.…”

The Temperature Forecast of Ship Propulsion Devices from Sensor Data

“…It is very difficult to select the most effective method from many prediction models, so we adopt the four most commonly used prediction models for getting the best one. In this study, four prediction models are employed: the autoregressive distributed lag time series model (ARDL) model, stepwise regression (SR) model, neural network (NN) model [22], and deep neural network (DNN) model, respectively.…”