You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This commit was created on GitHub.com and signed with GitHub’s verified signature.
Changes
Added primal and adjoint solvers for DARhoPimpleFoam 217f5e1. This solver is still in the beta state.
Changed the intermediate optimization folder format to 0.0001, 0.0002, etc. 5f2cc15
Allowed multi betaFI by using new names such as betaFINuTilda. 7570ed7
Added an option to use step-averaged states for steady-state solvers. 1b56851 This is for steady-state solvers with oscillating flow states due to separation.
Enabled using objective function's standard deviation as convergence criterion for steady-state solvers. d58cd15
Added the radial basis function for regressionModel. Also added the ReLU activation function. 35e0a4c
Allowed parallel field inversion cases with v3 run scripts. d0379b0
Allowed saving flow features to the disk for field inversion. 8afd2ab
Added TensorFlow callback functions for the regModel 304957a
Enabled using multiple regression models for computing multiple augmented fields in field inversion. be232cb
Allowed prescribing heat source for DAHeatTransferFoam. 8a9c93a
Added an option to use the latest flow fields instead of reading initial fields from the 0 folders for unsteady solvers. eb6d1b3
Dependency update
Updated the pyOFM version to 1.2.2 with new APIs for read/write fields.
Backward-incompatible changes
The naming convention for optimization intermediate folders has changed to 0.0001, 0.0002, etc. This will not change how we run the case, but it may create backward compatibility for some post-processing scripts that need to read the intermediate folders.
The augmented flow variable's name has changed from betaFI to more specific variable names, such as betaFINuTilda, betaFIOmega, etc. Users need to change betaFI in their runScript.py to the new naming convention.
The regressionModel dict has a new interface that supports multiple regression models. See an example from here
The tensorflow dict has a new interface that supports multiple regression model. See here