A new on-line combustion optimization approach for ultra-supercritical coal-fired boiler to improve boiler efficiency, reduce NOx emission and enhance operating safety

To take into account the economy, environment protection and operating safety of the boiler in the combustion optimization process, a new on-line combustion optimization approach for boiler is proposed. The historical combustion data collected from DCS of the coal-fired power plant is preprocessed at first. Then improved biogeography optimization-based long short-term memory neural network (IBBO-LSTM) and similarity measurement method are designed to construct the adaptive dynamic combustion model for boiler with boiler efficiency, NOx emission and the temperature of water wall as outputs respectively. After that improved non-dominated sorting genetic algorithm-II (INSGA-II) is designed to generate a series of boiler combustion optimization solutions under different running load offline, and improved multi-level fuzzy comprehensive evaluation (IDHGF) is designed to retain the combustion optimization solutions with higher running safety. Meanwhile, the retained optimization solutions are integrated into an optimization cases base. Thereafter, case-based reasoning based on safety enhancement mechanism (CBRSEM) is designed to achieve the online combustion optimization for boiler. Finally, to confirm the effectiveness of the combination of IBBO-LSTM, INSGA-II, IDHGF and CBRSEM, different online optimization methods (IBBO-LSTM-INSGA-II, IBBO-LSTM-INSGA-II-IDHGF, IBBO-LSTM-NSGA-II-DHGF-CBR, IBBO-LSTM-NSGA-II-IDHGF-CBR, IBBO-LSTM-NSGA-II-DHGF-CBRSEM, IBBO-LSTM-NSGA-II-IDHGF-CBRSEM, IBBO-LSTM-INSGA-II-DHGF-CBR, IBBO-LSTM-INSGA-II-IDHGF-CBR) are adopted to optimize a given combustion case. The proposed on-line combustion optimization approach for boiler received satisfied combustion optimization results that the growing for boiler efficiency was 0.653%, and the reduced concentration for NOx emission reached 22.891 mg/m(3), and the operating safety raised from 5.592 to 6.913. In conclusion, IBBO-LSTM-INSGA-II-IDHGF-CBRSEM can online offer the combustion optimization strategy to the boiler operators to improve boiler efficiency, reduce NOx emission and enhance the running safety of boiler, so that it is suitable for online application.

Automated summary

In this paper, a new online combustion optimization approach for boilers is proposed, which takes into account the economy, environment protection, and operating safety. It combines multiple algorithms and evaluation methods to improve boiler efficiency, reduce NOx emissions, and enhance operating safety. Experimental results demonstrate the effectiveness of the proposed method.