Open AccessPower unit load system modelling research based on BP neural network
Author(s) -
Ruicai Si,
Songhan Wang,
Xiwen Liu,
Li Jia,
Chi Zhou,
Baoju Li
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 - 10.1088/1755-1315/467/1/012101
Subject(s) - artificial neural network , inertia , unit (ring theory) , unit load , control engineering , power (physics) , control theory (sociology) , power system simulation , computer science , electric power system , coupling (piping) , thermal power station , variable (mathematics) , control (management) , engineering , mechanical engineering , artificial intelligence , mathematics , electrical engineering , mathematical analysis , physics , mathematics education , classical mechanics , quantum mechanics
A thermal power unit is a very complex non-linear multi-variable control object. It has the characteristics of strong coupling, large inertia, large delay and time-varying parameters. The traditional mathematical model is difficult to describe the non-linear characteristics among the parameters of the unit, and it is difficult to meet the control requirements of unit load, main steam pressure and so on. Based on the more advanced BP neural network modeling method, the mathematical model of the operation of 600MW thermal power unit under low, medium and high load is established by using the actual operation data on site. The method used is novel, and the simulation analysis shows that the model also has a strong engineering practice.