Research Area:  Metaheuristic Computing
In order to reduce the NOx emissions concentration of a circulation fluidized bed boiler, a modified teaching–learning-based optimization algorithm (MTLBO) is proposed, which introduces a new population group mechanism into the conventional teaching–learning based optimization algorithm. The MTLBO still has two phases: Teaching phase and Learning phase. In teaching phase, all students are divided into two groups based on the mean marks of the class, the two groups present different solution updating strategies, separately. In learning phase, all students are divided into two groups again, where the first group includes the top half of the students and the second group contains the remaining students. The two groups also have different solution updating strategies. Performance of the proposed MTLBO algorithm is evaluated by 14 unconstrained numerical functions. Compared with TLBO and other several state-of-the-art optimization algorithms, the results indicate that the MTLBO shows better solution quality and faster convergence speed. In addition, the tuned extreme learning machine by MTLBO is applied to establish the NOx emission model. Based on the established model, the MTLBO is used to optimize the operation conditions of a 330 MW circulation fluidized bed boiler for reducing the NOx emissions concentration. Experimental results reveal that the MTLBO is an effective tool for reducing the NOx emissions concentration.
Keywords:  
NOx emission
population group mechanism
teaching phase
learning phase
solution quality
faster convergence speed
Author(s) Name:  Yunpeng Ma, Xinxin Zhang, Jiancai Song, Lei Chen
Journal name:  Knowledge-Based Systems
Conferrence name:  
Publisher name:  Elsevier
DOI:  10.1016/j.knosys.2020.106599
Volume Information:  Volume 212
Paper Link:   https://www.sciencedirect.com/science/article/abs/pii/S0950705120307280