Hydrogen shaft furnace gas-solid coupling modeling method based on physical information neural network enhancement

Abstract

The invention provides a hydrogen shaft furnace gas-solid coupling modeling method based on physical information neural network enhancement. Aiming at an NS equation solving part required by gas phase motion modeling, enhancing a mixed function part of an SST k-omega turbulence model on the basis of a large amount of historical data of the hydrogen shaft furnace by using Neural ODE, and improving the solving precision of the SST k-omega turbulence model in different areas in the furnace; the method comprises the following steps: for a solid phase motion modeling part, calculating stress among solid particles by using a discrete element method (DEM), and modeling a motion equation of the solid particles by using a Newton's law; for the coupling modeling part, the solid-phase motion equation and the gas-phase motion equation are coupled based on a drag force model. According to the method, the complex fitting capability of the PINN is utilized, so that the turbulence models can be switched among different areas more accurately, and a more accurate result can be obtained for modeling of the gas-solid coupling motion in the hydrogen shaft furnace.

Images