LSSVM modeling for boiler combustion and denitrification integrated system based on adaptive GA variable selection

Guo Kaixuan, Huang Pingping, Hongwei Wang

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    The main cost of coal-fired power plants comes from coal consumption, and, in recent years, strict requirements for NOx emissions make denitrification an increasingly important part of operating costs. Therefore, the establishment of an effective integrated model of boiler combustion and denitrification is the basis for power plants economic optimization. In this paper, historical operation data are selected from the supervisory information system (SIS) of a 990MW thermal power plant. Combining the improved adaptive GA with the least squares support vector machine (LSSVM), the input variables are selected by the adaptive genetic algorithm (GA) to reduce the dimension and complexity of the model. The selected variables are used as the input of the LSSVM model and a GA-LSSVM model for a boiler combustion and denitrification integrated system is established. Comparing the model with the simple LSSVM model, the simulation results show that the complexity of integrated model can be effectively reduced by variable selection, the generalization ability of the model can be improved and the modeling time can be reduced. The integrated model can predict the SCR efficiency, SCR outlet NOx concentration and boiler efficiency accurately and rapidly.
    Original languageEnglish
    Title of host publication2017 IEEE 14th International Conference on e-Business Engineering (ICEBE)
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages269-274
    ISBN (Electronic)978-1-5386-1412-9
    ISBN (Print)978-1-5386-1413-6
    DOIs
    Publication statusPublished - 23 Nov 2017
    Event2017 IEEE 14th International Conference on e-Business Engineering -
    Duration: 4 Jun 20176 Nov 2017

    Conference

    Conference2017 IEEE 14th International Conference on e-Business Engineering
    Abbreviated titleICEBE
    Period4/06/176/11/17

    Fingerprint

    Dive into the research topics of 'LSSVM modeling for boiler combustion and denitrification integrated system based on adaptive GA variable selection'. Together they form a unique fingerprint.

    Cite this