SCR catalyst life prediction method based on BP neural network algorithm

Abstract

The invention relates to an SCR catalyst life prediction method based on a BP neural network algorithm, and belongs to the technical field of power plant denitration. According to the SCR catalyst life prediction method based on the BP neural network algorithm, data mining is carried out on DCS data of the denitration system, key indexes of catalyst activity are obtained, the BP neural network algorithm is used for more accurately predicting the change trend of the catalyst activity, a theoretical basis is provided for catalyst management of the denitration system, the prediction structure is accurate, and the prediction cost is low. Shutdown operation of a power plant denitration system is not needed, and normal work of a power plant is not affected.

Description

technical field [0001] The invention relates to a method for predicting the service life of an SCR catalyst based on a BP neural network algorithm, and belongs to the technical field of power plant denitrification. Background technique [0002] At present, more than 90% of coal-fired power plants in my country use selective catalytic reduction (SCR) for flue gas denitrification. SCR denitrification technology is mature and has a higher denitrification efficiency, reaching 80-95%. And nitrogen, will not cause harm to the environment. The principle of SCR denitrification is that under a certain temperature and the catalysis of the catalyst, the reducing agent such as ammonia gas reduces the NOx in the flue gas to nitrogen. [0003] As the core component of the SCR denitrification system, the catalyst's activity determines the denitrification performance of the denitrification system. Under complex and changeable working conditions, the catalyst activity will gradually decreas...

AI Technical Summary

Problems solved by technology

[0007] The input parameters of the catalyst activity prediction model need to be measured under laboratory conditions, which requires the denitrification system of the power plant to be shut down to complete; the variable for predicting catalyst activity is single, and the impact on catalyst activity is not comprehensively combined with flue gas parameters. Inaccurate activity prediction results

Images