Prediction of Coal Ash Flow Temperature Based on Gray Relational Analysis, Support Vector Regression and Genetic Algorithm

Abstract

Coal ash flow temperature significantly influences the operating conditions of entrained flow bed gasification. The relationship between the flow temperature of coal ash and its chemical composition remains uncertain despite being determined by it. To construct a reliable and accurate predictive method, machine learning models were used, and different support vector regression models were built to predict the flow temperature. The prediction results of the proposed gray relational analysis−genetic algorithm−support vector regression model can achieve high accuracy, with a root-mean-square error of 28.37, mean absolute error of 19.48K, and average deviation of 1.58%. Moreover, the prediction results of the proposed model are more accurate and efficient than those calculated by FactSage software, with a mean absolute error of 93.73K. This demonstrates the viability of applying the proposed machine learning model to predict the flow temperature of coal ash and its promising potential application in the area of coal chemical engineering.