Optimization study of denitrification of fully burned high alkali coal based on genetic algorithm

To address the challenge of reducing the concentration of nitrogen oxide emissions (NOx) from boilers, this study proposes a multi-objective optimization and prediction model for the sintering process based on a back-propagation neural network (BP) by collecting 25 days of NOX flow data and corresponding operating parameters from a typical 660MW coal-fired unit of Tebian Electric (TBEA). The BP neural network was optimized by combining with genetic algorithm to shorten the learning time and improve the prediction accuracy of the network model. The results show that the regression coefficient R-squared (R2) of the model for predicting NOx is greater than 0.9 in both the training set and the test set, which further realizes the intelligent diagnosis of the flue gas flow data of coal-fired units. This research not only reduces the comprehensive energy consumption of coal-fired units, but also provides key technical support for renewable energy systems, accelerates the deep integration of clean energy equipment and coal-fired systems, and helps to build a new type of power system with low energy consumption and high compatibility.