Open AccessImprovement of the setting method of back pressure setting value of direct air cooling system
Author(s) -
Jianyun Bai,
Qi Ren,
Xinyu Meng,
Jie Yin
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/467/1/012003
Subject(s) - particle swarm optimization , back pressure , range (aeronautics) , artificial neural network , computer science , process (computing) , field (mathematics) , simulation , control theory (sociology) , mathematical optimization , engineering , algorithm , mathematics , mechanical engineering , artificial intelligence , control (management) , pure mathematics , aerospace engineering , operating system
Because the setting method of back pressure of direct air cooling system is not reasonable, the fluctuation range of back pressure and load change are larger when the unit is peak shaving, which reduces the economic efficiency of the unit. In this paper, based on the optimal back pressure of field data calculation, particle swarm optimization (PSO) is used to optimize the modeling of BP neural network. The results show that, compared with the conventional BP model, the back pressure setting value predicted by the optimized model is more accurate, the model precision is higher, and it is more suitable for industrial field application. This method also provides some reference for the modeling of other objects in the process of thermal production.