A novel online combustion optimization method for boiler combining dynamic modeling, multi-objective optimization and improved case-based reasoning

Abstract

A new boiler combustion optimization method for online application is proposed based on data-driven and case-based reasoning principle. First, an improved constrained fuzzy association rule (ICFAR) is proposed to extract the boiler combustion rules from historical combustion data. Second, improved particle swarm optimization-based long short-term memory neutral network (IPSO-LSTM) is utilized to constructed an adaptive dynamic combustion model for boiler to adapt to time-variability of boiler combustion system. Whereafter, an improved multi-objective particle swarm optimization (IMOPSO) is designed to achieve the multi-objective combustion optimization of boiler with the goal of increasing boiler efficiency and decreasing NOx emission simultaneously. Case-based reasoning based on error correction mechanism (CBR_ECM) is applied to realize the boiler online combustion optimization. The effectiveness of proposed online combustion optimization technology for boiler is testified by applying it to an actual combustion process. The results illustrated that it could take less time to gain a series of excellent operating solutions based on proposed online combustion optimization method, the boiler efficiency increased by 0.082 % on average and the NOx emission reduced by 6.491 mg/m3 on average. Therefore, proposed online combustion optimization method for boiler has the ability to achieve the online combustion optimization of boiler.