Rule-Based Intelligent System for Variable Importance Measurement and Prediction of Ash Fusion Indexes

Abstract

Ash fusion temperatures [AFTs: initial deformation temperature (IDT), softening temperature (ST), and fluid temperature (FT)] are standard keys to estimate behavior of ash oxide for using coal and controlling the slag making at boilers. In this study, the modeling of AFTs based on ash oxide contents for 6537 U.S. coal samples have been investigated by a rule-based intelligent system (RBIS). Variable importance measurements (VIMs) of RBIS through the database indicated that Al2O3 contents in coal samples have the highest importance for prediction of AFTs. The RBIS model based on various rules was generated for predictions of IDT, ST, and FT. A comparison between RBIS and other typical predictive models [linear regression, genetic algorithm–neural network (GA–NN), and multilayer perceptron trained by back-propagation algorithm (MLP-BP)] was implemented to assess the capability of this purposed predictive model. Results indicated that RBIS can quite satisfactory predict AFTs, where R 2 for IDT, ST, and FT for the testing stage of models was over 0.82 and differences between actual and RBIS-predicted values for over 80% of data were less than 100 °C. These comprehensive results indicated that the RIBS method can be used for the industry sector to model AFT of coal samples and predict their fouling behavior before feeding them to boilers. Moreover, outcomes of this investigation are introducing RBIS as a powerful method for modeling other complicated problems in coal geology, fuel, and energy sectors.

An Interpretable Extreme Gradient Boosting Model to Predict Ash Fusion Temperatures

“…An AFTs measurement is generally expensive, time consuming and hard to repeat because it could generate an error during laboratory analysis [10,11]. Moreover, the ash fusion laboratory test has been questioned as it is more like a quantitative observation [12].…”

“…Another direction of research is the use of machine learning methods to predict AFTs [11,[20][21][22][23][24][25][26][27][28][29]. The most frequently techniques used were artificial neural networks (ANNs) and support vector machine (SVM).…”

Correlation between the Flow Temperature and Mineral Factor for Coal Ashes Based on FactSage Calculation

“…The accuracy of these correlations is relatively less because the AFT at a given condition is determined by mineral species and its content. , However, determination of the mineral composition of ash at a certain condition is complex and time-consuming. Fortunately, FactSage software provides a simple method to calculate the ash mineral composition and its variation with increasing temperature. , Although the prediction of AFT has been explored recently from different perspectives (e.g., liquid slag fraction or the temperature corresponding to a certain liquid fraction, , thermomechanical analysis, rule-based intelligent system model, liquid-phase content stage and its variation trend, and the diffraction intensity ratio of refractory minerals and fluxing minerals), the papers on the quantitative correlation between AFT and mineral compositions are relatively lacking, especially for the mineral composition calculated based on FactSage software. Thus, the aims of the paper are to establish the quantitative correlation between the FT and mineral composition (including liquid slag content) at a certain condition and to provide a simple way to predict the FT using FactSage software according to ash oxide composition in a reducing atmosphere and atmospheric pressure.…”

Research on semantic matching algorithm of BERT intelligent question answering system

“…On the other hand, the needs of document writers have become refined and personalized, and various industries have specific document formats and content expression requirements. Each web link is related to key words, but to get clear answer information, you need to open web links and browse web information to get answers, which is cumbersome and the answer hit rate needs to be improved 3 . Semantic matching technology in the field of natural language processing (NLP), as a close link between text representation and upper application, has extremely important significance and value, and has been widely studied and used 4 .…”