Open AccessDeep neural network based concentration prediction model of dry electrostatic precipitator with a modified PID optimizer
Author(s) -
Yao Fu,
Zhigang Su
Publication year - 2022
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2189/1/012018
Subject(s) - electrostatic precipitator , artificial neural network , pid controller , thermal power station , coal , process engineering , electric potential energy , power (physics) , automotive engineering , engineering , environmental science , simulation , computer science , waste management , control engineering , artificial intelligence , temperature control , physics , quantum mechanics
The dust control technology of thermal power units plays an important role in reducing the emission of air pollutants and improving the ecological environment. Electrostatic precipitator is not only an important equipment to reduce dust emission from coal-fired power plants, but also a high energy consuming equipment. Establishing the model of electrostatic precipitator is of great significance for energy saving of electrostatic precipitator and emission reduction of power plant. In this paper, a deep neural network based concentration prediction model of dry electrostatic precipitator of 600MW coal-fired unit is established and trained through running data. In the process of model training, an improved PID-based optimizer is used. The modified optimizer is proved to have the effect of accelerating convergence and improving accuracy in experiments. The modeling results demonstrate that the trained neural network model can effectively predict the outlet concentration of dry electrostatic precipitator under wide unit load.