Abstract: The combustion state of the boiler is an important part of the state assessment of the power plant, and its diagnosis has important practical significance. Through correlation analysis of on-site historical operating data, we searched for state parameters that are strongly related to the combustion state of the boiler, such as flame detection intensity, furnace pressure, total air volume, etc., and extracted characteristic quantities from them. A multi-level fuzzy comprehensive evaluation model based on state parameters was established using state parameters and their respective characteristic information to diagnose the combustion state. In the process of model establishment, in order to reduce the influence of subjective factors on the evaluation results, the CRITIC method was introduced to calculate the weight so as to make the evaluation results more objective and effective. Taking a 660 MW coal-fired unit as an example, the fuzzy comprehensive evaluation of the boiler combustion state based on state parameters was carried out. The results show that the method can objectively and effectively assess the combustion state of the boiler.