Control Strategy for Denitrification Efficiency of Coal-Fired Power Plant Based on Deep Reinforcement Learning

Jigao Fu,Hong Xiao,Junhao Zhou,Hao Wang

doi:10.1109/access.2020.2985233

Abstract

The optimal control of denitrification system in coal-fired power plants in China has recently received widespread attention. The accurate prediction of denitrification efficiency and formulate control strategy of denitrification efficiency can guide the control and operation of the denitrification system better. Meanwhile, it can achieve the effect of energy conservation and Nitrogen oxides (NOx) reduction. In this paper, we take a domestic 1000 MW unit as an example, consider each of the major factors that affect the denitrification efficiency of selective catalytic reduction (SCR). We put forward a deep reinforcement learning (DRL) model by combining the Long short-term memory (LSTM) model and the Asynchronous Advantage Actor - Critic algorithm (A3C). We first use the LSTM to build a prediction model for denitrification efficiency. We then use the DRL model to obtain a control strategy for SCR denitrification efficiency in coal-fired power plants. The experimental results demonstrate that the accuracy of denitrification efficiency prediction model we established is better than other machine learning models, reaching 91.7%. Our control strategy model is industrially feasible and universally applicable.

Highlights

As the industry continues to develop, Nitrogen oxides (NOx ) emissions are increasing continuously
1) We originally propose a deep reinforcement learning model that combines Long short-term memory (LSTM) and A3C algorithm, which can provide control strategies for the control of denitrification efficiency or related industrial applications
The analysis results demonstrate that the cumulative contribution rate of the 6 main components of the Boiler load, Inlet flue gas temperature, Inlet O2 mass concentration, Inlet NOx mass concentration, Ammonia/air mixer ammonia inlet regulator valve and Spray ammonia mass flow to the original data is greater than 90%, so we select the 6 main components as input to the LSTM model

Summary

Introduction

As the industry continues to develop, Nitrogen oxides (NOx ) emissions are increasing continuously. Acid rain has changed from sulfuric acid type to composite type of sulfuric acid and nitric acid [1]. NOx has gradually become the main source of gaseous pollution. For power station boilers burning pulverized coal, the pollutants’ NOx emissions are mainly Nitric oxide (NO) and Nitrogen dioxide (NO2), of which NO accounts for more than 90%, so NO and NO2 are generally referred to as NOx. In recent years, the government and research institutes have done a lot of research on the control of NOx pollution, they have developed many practical and efficient new technologies. According to the different control stages of nitrogen oxides during combustion, the emission reduction technologies are generally divided into before combustion, during combustion and flue gas denitrification after

Objectives

Results

Conclusion