Prediction of temperature distribution in a furnace using the incremental deep extreme learning machine

• Published in: PeerJ. Computer science Vol. 9; p. e1218
• Main Authors: Lv, Manli, Zhao, Jianping, Cao, Shengxian, Shen, Tao, Tang, Zhenhao
• Format: Journal Article
• Language: English
• Published: United States PeerJ. Ltd 17.02.2023; PeerJ, Inc; PeerJ Inc
• Subjects: Accuracy; Acoustics; Algorithms; Algorithms and Analysis of Algorithms; Artificial Intelligence; Artificial neural networks; Benchmarks; Boundary conditions; Carbon; CFD simulation; Cluster analysis; Clustering; Coal-fired power plants; Combustion; Computational fluid dynamics; Data Mining and Machine Learning; Fluid dynamics; IDELM algorithm; K-means clustering; Machine learning; Mathematical models; Methods; Neural networks; Partial differential equations; Prediction models; Radiation; Random sampling; Sampling methods; Simulation; Simulation methods; Temperature distribution; Vector quantization; Weather forecasting; CFD simulation; Temperature distribution; IDELM algorithm; K-means clustering
• ISSN: 2376-5992; 2376-5992
• DOI: 10.7717/PEERJ-CS.1218

In this article, a data-driven model based on the incremental deep extreme learning machine (IDELM) algorithm is proposed to predict the temperature distribution in the furnace. To this end, computational fluid dynamics (CFD) simulations are carried out first to get temperature distributions under typical working conditions. Based on the air distribution mode, the simulation results are divided into six subclasses. Then the K-means clustering method is applied to find out the benchmark working condition of each subclass. Moreover, the random sampling method is used to extract samples to reduce computational complexity. Modeling inputs are selected according to the CFD boundary conditions and combustion mechanisms, and data sets are reconstructed based on the increments of each actual working condition from the benchmark working condition. Finally, an IDBN-based prediction model is built in each subclass. The experimental results show that the IDBN-based model has a promising predictive ability with less than 11% symmetric mean absolute percentage error.