Case-based reasoning based on grey-relational theory for the optimization of boiler combustion systems

Boiler combustion optimization is an important method to improve the flexibility of thermal power units and ensures the stability of unit operation. However, time-variability of boiler combustion systems and time-consuming optimization methods pose great challenges for the use of boiler combustion optimization techniques because many optimization methods cannot be used online in practical engineering due to time constraints. In this paper, we propose a case-based reasoning optimization method based on grey-relational theory (GR-CBR) for online optimization of a boiler combustion system. After the introduction of the proposed algorithm, we discuss the potential of applying the proposed GR-CBR optimization method to a boiler combustion system; a case study of an existing fossil fuel power plant is conducted to demonstrate the feasibility of the proposed method. A least-squares support vector machine (LS-SVM) model of the boiler combustion process is established by using the real-time operation data of a 350-MW coal-based power plant. Based on the model, a non-linear global optimization algorithm is proposed to obtain the optimal case base and real-time data mining and online optimization are used to achieve efficient and stable boiler combustion optimization. The results of combining offline optimization with online querying show that this approach is suitable for online real-time combustion optimization, and provides support for power plant operators for optimization and condition monitoring to improve boiler efficiency, reduce NOx emissions, and ensure stable and efficient operation of the power system.