In this paper, taking improving the flexibility of peak load regulation and frequency regulation of thermal power units and promoting the consumption of renewable energy sources as the target, the combustion stability and economy of a certain thermal power unit during operation is studied. The adaptive genetic algorithm is adopted to optimize the kernel function parameters and normalization parameters, and a least square support vector machine (LSSVM) boiler combustion process model is established. On the basis of the established LSSVM model, an offline optimized case base is established using the adaptive genetic algorithm. Then, from the perspective of facilitating engineering application, a casebased reasoning (CBR) optimization method is proposed. In consideration of subjective and objective factors, the genetic algorithm is used to optimize the feature weight of CBR, which improves the retrieval accuracy and adaptively retrieves the case matching with the target case from the huge case base. The application of the CBR adaptive optimization algorithm ensures the stable combustion of the unit, and at the same time, considers the boiler combustion efficiency and the concentration of NOx emission. This algorithm reasonably gives the opening instructions of the secondary and tertiary valve baffles and the fixed value of oxygen, and realizes the economic combustion of the boiler. The system was applied to a certain 350MW coalfired generation unit, which simplifies the process of optimization calculation, shortens the optimization time and has high stability. The system is suitable for online realtime optimization.