class SIMP[source]
SIMP(criterion:dl4to.criteria.Criterion,p:float=3.0,n_iterations:int=50,verbose:bool=True,lr:float=0.03,binarizer_steepening_factor:float=1.0,density_representer:dl4to.density_representers.DensityRepresenter=FilteringDensityRepresenter(),return_intermediate_solutions:bool=False) ::TopoSolver
A topo solver that performs topology optimization with the "Solid Isotropic Material with Penalization" (SIMP) method.
| Type | Default | Details | |
|---|---|---|---|
criterion |
dl4to.criteria.Criterion |
The objective function that should be optimized for in the optimization process. | |
p |
float |
3.0 |
The SIMP exponent to discourage non-binary densities. The default value is p=3, which is the standard value in the literature. |
n_iterations |
int |
50 |
The number of SIMP iterations that should be performed. |
verbose |
bool |
True |
Whether to give the user feedback on the current status of the optimization. |
lr |
float |
0.03 |
The learning rate of the torch.optim.Adam optimizer. |
binarizer_steepening_factor |
float |
1.0 |
The factor at which the binarizer should be steepened in each iteration. E.g.,a value of 1.1 corresponds to a steepening of 10% per iteration. |
density_representer |
dl4to.density_representers.DensityRepresenter |
FilteringDensityRepresenter() |
The density representer that is used for the latent density representation. The density representer also performs the projection, smoothing and filtering. |
return_intermediate_solutions |
bool |
False |
Whether intermediate SIMP solutions should be returned or only the final solution of the optimization process. |