SCR system NOx generation and ammonia demand prediction method and system and storage medium

Abstract

The invention discloses an SCR system NOx generation and ammonia demand prediction method and system and a storage medium. Three types of machine learning models for SCR system inlet NOx concentration prediction are obtained through training according to historical operation data of an SCR system, corresponding inlet NOx concentration prediction results are obtained, then a mathematical statistical method is used for analysis, decision rules are agreed, and an inlet NOx concentration prediction output result to be adopted is obtained; in combination with other real-time on-line monitoring operation data, the ammonia demand predicted quantity Q' of a urea hydrolysis system is obtained; and the inlet NOx concentration prediction output result is subjected to precision judgment in real time, and online self-updating rules of the three types of machine learning models are formulated. According to the method and the system, the operation problem of poor variable working condition following characteristics caused by large response slag time in the urea hydrolysis ammonia production process is solved, and precise ammonia spraying of the SCR system under a urea hydrolysis ammonia supply system is facilitated.

Description

technical field [0001] The invention relates to a NOx generation and ammonia demand prediction method, system and storage medium of an SCR system, and belongs to the technical field of flue gas denitrification in thermal power engineering thermal power plants. Background technique [0002] Selective catalytic reduction (SCR) flue gas denitrification technology is widely used in the effective removal of nitrogen oxides (NOx) in domestic thermal power plants due to its advantages of reliable technology and high denitrification efficiency. The preparation methods of SCR denitrification reducing agent NH3 can be roughly divided into Liquid ammonia method and urea decomposition method. The liquid ammonia method system is relatively mature, but according to the relevant requirements of GB18218-2009, a major hazard source identification standard for hazardous chemicals, liquid ammonia storage tanks exceeding 10 tons are already a major hazard source, and there is a relatively large...

AI Technical Summary

Problems solved by technology

In addition, the actual operation of the SCR system will be affected by factors such as instrument measurement characteristics, coal quality, boiler load, and combustion conditions, showing characteristics such as nonlinearity, large hysteresis, and multi-parameter coupling effects, especially the NOx concentration at the inlet of the SCR system. hysteresis

Therefore, the coal-fired power station SCR system equipped with urea hydrolysis system generally has a series of operational problems such as poor followability of variable working conditions of the urea hydrolysis system, unstable automatic operation of ammonia injection, and large ammonia consumption of the system, which urgently need to be optimized and solved

Images